I4SS19390 - Greenwich Academic Literature Archive (GALA)

H oo (1Rýii TP

"No Promised Land"

I4SS19390

History, Historiography &-the Origins of the Gypsies

submitted by Adrian Richard Nathanael Marsh in partial fulfilment of

the requirements for the degree of Doctor of Philosophy awarded by the University

of Greenwich London

November 2008

ttK

r

History, Historiography & the Origins of the Gypsies

iii. Acknowledgements

COMPLETING THIS BOOK HAS REQUIRED the sustained and collective engagement of many people,

and the support of many more during the process of writing. My thanks must go primarily to my

supervisors, Professor Thomas A. Acton and Professor Emeritus Angela V. John of the University of Greenwich in London, who have gone a great deal beyond the formal relationship between

candidate and supervisors to become mentors, friends and ̀ family'. I must also thank Dr Wendy

Bracewell and Professor David Kirby who first set me on the path to `modern historiography' and introduced me to the investigation of narratives of ethnicity, identity and nationalism in history and to questioning the writing of history itself. To Dr Peter Siani-Davies, Professor Norman Davies,

Professor Geoffrey Hoskins, Dr Martin Rady, Dr Roger Bartlett, and Caroline Newlove at SSEES,

Dr Ben Fortna at SOAS and Dr Kate Fleet of the Skilliter Centre for Ottoman Studies Cambridge

University, I would also like to extend my thanks for their patient instruction of me as an

undergraduate and post-graduate student, and their encouragement of the various, and sometimes

unusual paths that my inquiries took during the years I studied with them.

Many other eminent scholars and researchers have informed my work and encouraged me; Professor Yaron Matras at Manchester University, Dr Norma Montesino at Malmö Högskola (at the time), Dr Christina Rodell-Olgas at Södertörns Högskola, Dr Hans Lejdegard at Uppsala University, Dr Karin Adahl, Director at the Swedish Research Institute in Istanbul (SRII), Dr György Peteri and Professor Birgit Sawyer at Trondheim University, Professor Elena Marushiakova

and Professor Vesselin Popov at the Sofia Studii Romani, Professor Barbara Harrell-Bond at the American University in Cairo, Dr Karl Nylander (former Director at the Swedish Research Centre in Rome), Professor Alan Duben, Dr Ayhan Kaya and Nese Erdilek at Istanbul Bilgi University, Dr Emine Onaran Incirlioglu at Bilkent University Ankara, Dr Suat Kolukirik at Ege University, Izmir, Dr Sonia Tamar Seeman at the University of Texas at Austin (with whom I was privileged to

conduct fieldwork during summer 2007) all of whom I would thank and express my gratitude for their patient understanding, support and the opportunity to present my work and aspects of it at various points. My thanks also go to my examiners, Professors Ian Hancock and John Dunne for

clarifications during the exam.

The research that I have undertaken during the course of writing this book has been financially

supported by the Arts & Humanities Research Council, (in the UK), the Consulate General of Sweden in Istanbul (and in particular, the General Consul Dr Ingmar Karlsson and the Consul

4


Annika Svahnström), the Open Society Assistance Foundation in Turkey, the British Council in

Turkey, the Council of Europe and the University of Greenwich in London. I am very grateful for

their support in carrying out the various projects that, whilst not all directly connected to this book

have nevertheless informed its development and outcomes.

All scholarship, and historical research in particular relies upon access to libraries and I would like

to thank the Director and staff of the SRII library (where much of this book was written), Ms. Eva

Nylander at Lund University Library, Serdar Katipoglu at Istanbul Bilgi University Library, the

librarians of the American University in Cairo, Trondheim University Library, Malmö Högskola

Library, the British Library in London and the National Library of Scotland in Edinburgh. Their

assistance and technical advice has been of enormous help, particularly as I have been a visitor to

almost all of these.

My thanks to family and friends who have encouraged, supported, listened to, `put up with' and

sometimes cajoled me into continuing when I have been most despondent can only reflect here the

debt I truly owe them; thank you Neil Darby, Stephen Spencer, Özhan Onder, Bertil Videt, Dirk

Nieuwboer, Hege Irene-Markussen. To Ingeborg Stensrud, Fred and Maja Taikon, Karl and Eva

Nylander, Johnny Strand and Baiba Veldre, Belinda Acton, Paul and Felicity Bonel, all of whom have provided sustenance and much-needed comfort thank you. In particular Johnny and Baiba,

Karl and Eva were very kind whilst visiting Sweden, Belinda Acton, Paul and Felicity Bonel in

London, providing places to stay often for extended periods. My family, Keith and Clare Marsh,

Valerie Marsh, Nichola Scale, Joanna Marsh-Willis (who proof-read this text), Laura Marsh and Rosie Marsh have been and continue to be the `safe haven' from which I have set out on these journeys, whilst Nezihe, Faruk, Muhterem, Gamze and Ali Kula have made me one of their own,

cok tqekkür ederim. To Arzu Olcayli-Marsh, seni seviyorum ackim birtanem.

This book has been written during a period that has begun to see the development of a consolidated Romani Studies in Turkey and to those who have been part of this I would also extend my gratitude. To the friends who set up the original Istanbul (later International) Romani Studies Network, Ana Opriýan, Dr Udo Mischek, Dr Mustafa Özünal MD and Dirk Nieuwboer, thank you. I would also thank Elin Strand for her contribution to our work together in London, Stockholm, Malmo[], Istanbul, Sofia, Cyprus, Cairo and in general during 2002-2005.

Finally, the loss of Dr Leslie Collins and Professor Lindsey Hughes of SSEES, both much too young, deprives me of the opportunity to thank them as teachers, mentors and much-loved friends.

5


Leslie taught me to see the Ottoman Empire as an inheritor of Byzantium, and challenged me to

think beyond the overly-schematic and reductive notions such as ̀ the Turks', `Islamization in the

Balkans', `the Asiatic mode of production' and to question every assumption made by others and

myself. He also shared his vast knowledge about central Asian nomadic peoples and their impact

upon Europe, his love of 18th century English poetry (Samuel Daniel in particular), jazz from the 1950's and 1960's and was the only lecturer I knew who would demonstrate an Avar cavalry charge,

at full volume and speed in the lecture room. Lindsey taught me to understand that history can be

`read' in icons, buildings, statuary, city plans and laundry lists and it was with her that I first found

the Gypsies in history (the Skomorokhi or minstrels of Muscovy, many of whom were Roma). More

than any others, their encouragement and their belief in me enabled me to achieve what I had never thought possible and they remain, as they always will the inspiration behind this and all my

academic work. My hope is that they would have been pleased with it...

6


iv. Abstract

THIS BOOK EXAMINES THE QUESTIONS of how Gypsy ethnicity, identity and history are interlinked in

the context of examining various contested narratives of origins and migration. The text is itself a

series of narratives and counter-narratives that engage in a self-critical, deconstructive analysis of

the underlying assumptions hitherto presented in many, if not most of the previous scholarship

regarding the origins and identity of the Gypsies, with particular focus on the contextual and

radically contingent nature of all such texts. As such, the primary examination is an historiographical and theoretical consideration of the questions surrounding Gypsy ethnicity and identity, as they are embedded in competing versions of historical discourses claiming authenticity

and authority.

The dissertation also considers to what extent the production of historical knowledge - contested

and contingent as it must be - is affected by those who produce it from within and without the Gypsy community or communities themselves. The construction of historical narratives of journey,

undertaken by Gypsies, Roma, Travellers and Sinti themselves, is a relatively recent phenomenon in

academia and this dissertation reviews this development in a survey of literatures as they

particularly relate to debates on origin, ethnicity and identities. Most especially, this survey examines the production of literatures in Turkish scholarship, as related to the underlying conception of the book arguing for a re-examination of Romani historiography from east to west, rather than the `traditional' Orientalist and Europecentric perspectives deployed by much of the

previous scholarship. Aside from few monographs, the history of the Gypsies in the Ottoman Empire and Turkish Republic has received scant attention to date.

Moreover, the dissertation focuses upon the Turkish lands to argue that the historical experiences of Gypsies in this region are of critical importance in understanding the development of both European Romani histories and in acknowledging the flawed basis for the universalist conceptions of European Roma identity and political mobilisation, as they are now articulated. In this context, the importance of Islam in the origins and history of the Gypsies is stressed in the text.

This theoretical framework underlies the interweaving narratives that make up the latter sections of the text, a reconsideration of the sources for early Gypsy history that posits an alternative narrative. Whilst openly acknowledging its contextual and contingent nature, I present this as a plausible solution to some of the problems of origins, especially for the Dom Gypsies of the Arab, Persian

and Turkish lands and the Lom of Turkey, Iran and central Asia. The recapitulation of the linguistic

7


arguments regarding Romani origins I undertake in this book from a historiographical perspective,

challenging the preponderant use of historical `evidence' as merely a background to certain linguistic assertions, in an effort to `ground' these in historical contexts that support such

argumentation, before excising that which appears insupportable.

8


V. CONTENTS

i. Title

ii. Declaration

iii. Acknowledgements

iv. Abstract

v. Contents

vi. Glossary of particular transliterated words and letters

vii. Byzantine, Armenian, Muslim dating: a guide Preface

Introduction

1. Historiography

1.1 Historiography: History, Ethnicity & Identity

1.2 The "Grand Narrative" in Romani historiography 1.3 Stout hearts and sturdy footwear: the Gypsy Lore Society

1.4 Gypsies, history and the problem of time

1.5 Language-change, loan-words and war 2. Who are the Gypsies?

2.1 Who are the Gypsies? Identity and influences in Romani Studies 2.2 "... the strumming of their silken bows... " the legend of Bahram Gür... 2.3 The origins of the Romani connection 2.4 The interpolations of Colonel John Staples Harriott

2.5 al-Zutt and the Lüli

2.6 "Strong and valiant men"; the Hephthalites

2.7 Harriott, history and historiography

3. India, Islam & the Dom

3.1 The origins of Domara in the Umayyad Khalif a

3.2 Arabic as a military koine

3.3 al-Sind, al-Hind and Rädjpütana in 8th century 3.4 The Umayyad and `Abbdsid Caliphates: the Hindus into Dom

4. Ghazni! & the origins of the Rom

4.1 Koines and cavalry charges

4.2 Ghuläms, rnawlil and `abld; the enslavement of Hindus

3

4

7

9

11

11

12

14

26

33

45

60

70

74

84

93

95

100

103

108

109

115

118

124

128

134

9


4.3 "Speaking to the scribes, talking to the troops... " 141

4.4 State and structure: palace and chancellery in the ghazi emirate 146

4.5 The ideology, military organisation of the Ghaznävid state and the Hindus 150

4.6 The Ghaznävid campaigns in India: ghazawät and razzias 160

4.7 The city as theatre; Ghaznävid culture & literature 167

5. The Saldjükids

5.1 The Saldjükids, Armenia and Byzantium 169

5.2 Mas'iid b. Mahmüd and the fragmentation of Ghaznävid authority 172

5.3 "... no sultäns greater ... than the kings of the house of Saljuq... " 178

5.4 The irruption of the Saldjüks into Armenia 185

6. The Byzantine Empire; the origins of the `Egyptians'

6.1 The collapse of Byzantine Anatolia 1071 CE - 1176 191 6.2 Taxation, Gypsies and status in the Byzantine Empire 198 6.3 Magic and Romitoi in Byzantium; the construction of `Gypsy' identity 204

7. The Ottoman Empire

7.1 Ottomans and Gypsies 213

7.2 Gypsies in the late Ottoman Empire 218

8. Conclusion

8.1 Time, narratives of journey and the writing of Gypsy history 233

Bibliography

10


vi. Glossary

of particular terms, transliterated words, letters and their sounds, as they are used in the text

ì' similar to the English sound ̀ er'

ì' as in èe' in èasy'

69' as in `ch' in `church'

`c' as in `j' in `jam'

as in `sh' in `shout'

as in àh' in `chart' and ̀ castle'

as a silent or `soft' letter `g' that lengthens the preceding vowel

following the usage of the Encyclopaedia of Islam:

Àmir' rather than Èmir'

`Sultän' rather than Sultan

`Ghaznävid' rather than Ghaznavid/Ghaznawid

`Saldjük' rather than Selcuk

`Jhät' rather than `Jat'

`Hindu' rather than `Hindu'

`Romani' rather than `Romani'

`Doman' rather than `Domari'

`Zutt' rather than `Zott'

`Khalif a': the trusteeship of the umma, entrusted to the successors to the Blessed Prophet

Muhammad, the first four of whom - Abu Bakr (632-34 CE), Omar (634-44 CE), Othman

(644-56 CE) and All (656-61 CE) - are known as ̀ Rightly Guided' Khalifs

`ghazwära' holy war, prosecuted by `ghazi warriors' `dar Ul-Islam' the House (or Abode) of Islam, as opposed to `dar ül-Harb' the House of War

(i. e. non-Muslim territories)

vii. Byzantine, Armenian, Muslim dating: a guide BCE/CE before common era/common era (previously BC/AD)

The Byzantine Empire dating system "since the year of Creation"; i. e. 1000 CE = 6,509 YC

The Armenian Era begins 554 CE; i. e. 1000 CE = 449 AE

Islamic dating begins with al-Hijra (622 CE); i. e. 1000 CE = 390 AH

11


Preface

"The only good is knowledge, and the only evil is ignorance... " Herodotus 'Histories'(] 862)

THE I OT" AND 11TH CENTURIES were a period of crucial importance in the history of the Gypsy

peoples. In this book I will argue that over approximately 200 years (from ca. 977 to ca. 1200 CE)

the captive Hindü soldiers and servitors that were incorporated into the Ghzanävid Empire of

Mahmüd ibn Sebüktigin developed a lingua franca that lies at the basis of modern Romanes and

Urdu, before being thrust into a westward trajectory by the defeat of the Ghazndvids at Dandängän

in May 1040 CE (see Bosworth, 1973: 241-68), to emerge as the Ègyptians' of the Byzantine

Empire at the close of the 11`h century (see Fraser, 1992: 45-59). Their identity came as a result of a

series of confluences that brought interest in magic in Byzantium to a particular height at the time

of their earliest arrival, with the collapse of eastern Asia Minor and the loss of these lands to the

Empire bringing insecurity and the need for reassurance that enabled the Ègyptians' to exploit an

opportunity that came to define who they were, as fortune-tellers and diviners, agents of the

èxotika' (Gr. See Tomkinson, 2004). Such an identity and its relationship to notions of magic and

sorcery have lain at the basis for perceptions about Gypsies into modern times. This book attempts

to examine the circumstances and conditions in which these developments took place.

This book also examines the question of the presence of Gypsies in the Ottoman commonwealth as

the successor to the Byzantine Empire, and the relative absence of critical scholarship regarding

them therein. The economic and social position of the Gypsies in the Ottoman Empire underwent

significant changes during the period of modernisation and reform known as the Tanzimat, that

brought conceptions about them into the orbit of wider European notions concerning the Gypsies,

and was part of the Orientalisation of the Ottomans, with which the Ottomans engaged themselves

particularly under the renascent Kahlif'a ideology of Abdul Hamid II (1878-1908 CE), as regards

those they perceived as irreformable and backward, such as the Kurds, the Arabs, the Druze, the

Maronites and the Gypsies. This process was an aspect of the Ottoman attempt to resist the

discourses of colonialism and imperialism and establish a coeval model of Islamic civilisation, by

subjecting others to the process of `nesting Orientalisms' (Bakic-Hayden, 1995).

These notions were at the core of an original conception of this book as a work of history, a conception that has changed dramatically over the course of its production. As importantly this

book is now also a contribution to the debate about the writing of Romani history or histories and

suggests that the concerns of mainstream historiography over the past twenty-five years or so, have

12


largely been ignored by Romani Studies scholars who have not confronted the implications of the

questions raised by the work of Hayden White and others regarding the pursuit of history,

questioning "the content and the form" or the deconstruction of the `grand narrative' (White,

1987: 26-57). I question the underlying tropes that have been consistently used in the writing of Romani history and raise the questions of how and who this has been written for. The heart of this

book is now an examination of how narratives of the origin of the Gypsies relate to different

historiographical agendas. In passing, this throws light upon how we can develop a transcendent

rational agenda, which can develop a progressive humanitarian perspective because it problematises the explanations of inequalities of power and oppression, rather than taking them for granted as

nationalist historiographies (and Romani nationalisms) do. Inevitably this means the history of the Gypsies, as it is presented in the book is cast in terms of a succession of historiographies responding to the historical succession of agendas.

Since all these historiographies, except the anti-history of synchronic anthropology are built around accounts of origin (even those that purport to deconstruct such accounts), there is a particular emphasis on the exegesis of the 7`h -13th centuries drawing upon a range of area studies. I have

undertaken these in the pursuit of understanding particular problems that relate to the debate around origins, including the Dom Gypsies of the Turkish, Persian and Arab lands as they are frequently ignored or cursorily treated in most Romani Studies literature. I also attempt to posit an explanation for the division between the Lom and the Rom, in the catastrophic collapse of the Baghratid Armenian kingdoms in 1064 CE at the hands of the Saldjükids. I conclude with a summary of the

main points of the book and the arguments it contains.

I hope that this study will likewise serve as an introduction to some of the wider debates in history

and historiography, and how these might relate to the writing of Romani histories. I also hope that the examination of Gypsies in the Islamic lands, in a historical and historiographical context, will provide an introduction to groups of Gypsies inhabiting what has been, and to some extent remains terra incognita in the wider discipline.

13


Introduction

"Perhaps what I'm saying is not true, but it may be prophetic... " Jorge Luis Borges 'Theme of the Traitor and the Hero'

in 'Fictions' (1962)

THIS BOOK BEGAN AS AN ATTEMPT TO INVESTIGATE the social and economic position of Gypsies in

the late Ottoman Empire (1789 CE - 1923 CE). The principal approach was one that I had absorbed

from reading the secondary material regarding Gypsies, my own experience of living with and

amongst Gypsy, Roma and Traveller peoples and applying the knowledge acquired from study of

the Ottoman Empire and eastern European history at undergraduate and master's degree levels, as a

student at the School of Slavonic & East European Studies and the School of Oriental and African

Studies, both then part of the federated University of London. I had envisaged that an attempt to

combine a thoroughgoing analysis of demographic and statistical materials together with what

might be regarded as a `traditional' historical approach relying upon periodisation and a strict

chronology, would bring the `story' of the Gypsies from the least well documented parts of their

history to light, in an informative and useful addendum to the body of material examining the

Gypsy, Roma and Traveller peoples in the context of mediaeval and early modern Europe. In this

frame of mind, questions of origins, ethnicity and identity were best left to sociologists, ethnologists

and anthropologists, and the matter of Romani, Domara and Lomavren linguistics was one that be

effectively dealt with by experts far more knowledgeable than I.

This entailed the questioning of sources that by and large, were already in the public domain and

translated into modern Turkish or English, by scholars and historians, such as Kemal H. Karpat's

Ottoman Population 1830-191: Demographic & Social Characteristics (1985), Justin McCarthy's

The Arab World, Turkey & the Balkans 1878-1914: a handbook of historical statistics (1982), and

Stanford J. Shaw's seminal work on Ottoman populations (1978). The economic picture was one

that I hoped to introduce through the analysis of Ottoman Gypsy occupations, in the context of

referring to the work of Michael Palairet (1997), Donald Quataert (1994), and other economic historians working in the field, most especially of course, Halil Inalcik (1994). The culture of the

late Ottoman Empire, its views about the minorities living within its territories and the particular development of Ottoman Orientalism, were something that I had intended to portray and critique

using the work of Edward Said (1978), Cemal Kafadar (1995), Usaama Makdisi (2002), Ash

cirakman (2002), and Gabriel Pitterberg (2003), amongst others. The overall perspective that I had

initially formulated (following Angus M. Fraser's observation that in the Ottoman Empire "they

[the Gypsies]... were left pretty much unmolested by western European standards"; 1995: 175) was

14


one that suggested the Ottoman Empire had been a place of some degree of tolerance, or at least

benign indifference towards the Gypsies in what has been termed "the classical period" (Inalcik,

1973), and this had been undermined by increasing European influence throughout the 19"' and

early 20th centuries by pervasive notions about Gypsies that entered Ottoman discourse, an aspect of

the debate about the impact of European influence upon the lands of the Sultans and the question of

reforms. The numbers of European and most especially western European folklorists and

ethnologists that had come to the Ottoman lands in the later imperial period, had carried with them

the seeds of a scientific racism that had imbued much of their own work, framing the kinds of

questions that were conceived of and the perspective through which the Ottoman Gypsies became

viewed (see Sardar, 1977 for a discussion regarding the impact of scientific thinking from Europe

upon the Muslim world).

This Orientalist paradigm had come to be associated with a positivist school of sociology that had

dominated Turkish research (and to some extent, still deeply influences it; see Sardar, 1979: 668-70)

and analysis in the social sciences. This, combined with an increasingly negative view regarding all

minorities in the late Ottoman commonwealth (most especially the Christian millets; see McCarthy,

1997: 204-209), led to a significant deterioration of the social position of Ottoman Gypsies, I

thought to conclude. The notions of propriety, always a point of contestation during upsurges of

religiosity during the rule of various sultans, also influenced the view of Gypsies as indifferent

Muslims and Christians, mostly nominal in their adherence to Islam or to Christianity for that

matter, under the revivalist Khalif a ideology of Abdul Hamid II (1876-1908) in particular (Karpat,

2001; Stoddard, 1922: 37-75; Blunt, 1882: 48-90). At points the Sultan actively intervened in the

prohibition of Ottoman Gypsy dancers touring abroad for example, on the grounds that it would be

a shameful exhibition that would encourage negative European views of the Empire (Derengil,

1998: 62). Thus Gypsies were alienated in a way that reflected European notions about them in the

context of an Islamic religious world-view, and embedded in a discourse about definitions of the

`Other', in the developing Ottoman Orientalism, coterminous with prejudices about the Arabs,

Kurds, Druze, Maronites and various other minorities seen to be alien, different, disloyal and

ultimately irreformable. This was a significant change from the so-called classical period, pre- Tanzimat (the period of reform that begins in the 1830's and ends with the accession of Abdul

Hamid II, though the efforts of other sultans, in particular Selim III in the 1790's should also be

seen as part of the programme; see Shaw, 1971), and profoundly European influenced Empire,

where Gypsies had a place in the complex ethnic and religious mosaic that was Ottoman society, or

15


so I had conceived of arguing.

In order to establish the context for Gypsy populations in the late 19th century, it was necessary to

explain their historical presence and arrival in the Ottoman lands. The most recent scholarship

relied almost solely upon information gleaned from Byzantine records by George Soulis in his

article from over forty years ago (1961: 141-165), consistently referred to by authors since

(Marushiakova & Popov, 2001: 11-21; White, 2000; for examples), with reference to very few other

primary sources. The documents cited by Soulis and others following him seemed to describe a

picture of Romani presence in the Byzantine lands from as early as c. 1056 CE, but aside from one

rather cryptic indication in a commentary upon an obscure ruling at the Council of Trullo, itself in

692 CE, little else elaborated upon this date. As to any indication as to how the ancestors of Romani

people may have reached the Byzantine capital, the suggestion was only slightly made that their

passage was facilitated or hastened by the advent of the Saldjiiks in the region, beginning about the

Armenian year 467 AE/1018-19CE, though little historical argument had been delivered to support

such a notion. Alternatively, the predominant narrative suggested that the migration of those we call Gypsies could be traced through references to the Persian lands and the groups known as Lüri or

LUl , following Col. John Staples Harriott of the Bengal Infantry (1830: 518-558; see Minorsky,

1986: 816-819; Grierson, 1889: 71-6). Despite a significant lexicon that is apparently derived from

Armenian in modern Romani, Domari or Lomavren, little investigation had been made to suggest how and where these ancestral Romani peoples may have come into contact with the language or

people of Armenia, or under what circumstances such a lexicon could have been acquired, other

than suggestions regarding a Romani sojourn in the Armenian lands (Dr. Vardan Voskanian has

developed the most thorough scholarship regarding Armenian Gypsies and what is described as Lomavren, or a ̀ secret language' to date; see 2003: 169-80). Likewise, the context for the

acquisition of numerous Greek terms in modern dialects of Romani has not been especially investigated by scholars, but asserted without a more detailed suggestion of the context of time and

place. Though perhaps the consideration of this context could only ever be speculative, it might still

serve to illuminate the possibilities surrounding how Gypsy identities came into being, and under

what pressures or factors there might have been at play.

The earlier Indian connection has been much vaunted, both as part of the continuing scholarship

associated with the political emancipation of Romani and other Gypsy peoples, and as a topic of inquiry in and of itself. The contested nature of an Indian connection notwithstanding (see Mayall,

2004: 119-125), the majority of scholarly opinions and the evidence of linguistic research does

16


accept this as primary in the discussion of the origins of the Romani, Domart and Lomavren-

speaking peoples, and it has been incorporated into the notions of self-identity for a large number of

Gypsy groups (see Matras, 2005: 53 - 78). What is less clear is the process by which the ancestors of

modem Gypsy communities left India, or were forced to leave and how they came to be a migrant

communities in territories far removed from any "homeland". The discussion of the role of the 11th

and 12th century Islamic Turko-Afghan dynasty of the Ghazävids in early Romani history has

proved the departure point for much of this book (Hancock, 2006: 81-2; 2002: 10,13-14 and Fraser,

1995: 28) regarding the state, organisation and campaigns into Hindü and Ishma l India. The pivotal

role of the Sultan Mahmüd of Ghaznä (998-1030 CE), has been examined closely here, as I have

attempted to draw out a narrative that looks in more detail at the circumstances whereby groups of

Hindus and Ishmailis became part of this Islamic power-state, and how they may have been

identifiable with the ancestors of the Romani people, at least in part. Other scholars have chosen to

focus upon what they see as the more negative aspects of Mahmüd and the Ghazänivd impact upon

India; indeed, the more `Hindutva' of the contributors to the Roma Virtual Network insist upon the

calamitous effects of these Ghaznävid incursions ("Hindutva is not a word but a history"; see

Savarkar, 1949), and in effect follow those who ascribe to the pre-eminence of Hindu civilisations in the sub-continent and refute the so-called Aryan invasions in favour of an indigenous origin for

these (Thapar, 2000b: 15). They are often deeply antithetical to Muslim and Christian influences and impacts upon the history of India and the region (Thapar, 2000b: 16), and demand that these be

treated as "black days" in Rroma history (see Roma Virtual Network, 2007 and responses, 20th

December). Some reconstructed narratives of early Roma history have been profoundly influenced

by this highly politicised notion (see Courthiade's 2007, "Short Chronology of Rrom's History",

posted to the Roma India group, 20th December), and the growth of modem Hindu nationalism in

India since 1948 - in many ways a parallel to the development of Rroma nationalism and the

International Romani Union (Rroma in this context is a term that reflects a particular identification

that might be said to emphatically assert an ethno-linguistic notion of identity over cultural or other

attributes; see W. R. Rishi, 1995 for a counter argument to this use by "neo-linguists"). The

questions surrounding the impact of Hindutva ideology upon Romani identity and politics has not been part of the discussion of origins, to the best of my knowledge, but the implications of this are I

would argue here, important in the construction of Romani identity and ethnicity.

The investigation of the Indian `origins' of the Gypsy peoples needed more than the replication of

rhetoric asserting such, particularly in the wake of challenges to this notion from the post-modem

17


scholarship of what has become known as the `Dutch school' (Willems, Cottaar, Lucassen, 1998;

van de Port, 1998; Duijzings, 1997). These and others (see Belton, 2005, Mayall, 2004) have sought

to `de-exoticise' the Gypsies through arguing for a complex understanding of the processes of

identity formation that challenges the ethnic basis of much of Romani Studies scholarship; a

position that seemed confirmed by my experience of carrying out field research amongst the

Zabaleen and Dom communities of Cairo in 2000. Here a group of hereditary, Coptic Christian

garbage collectors that inter-married with the nearby Dom communities appeared to be emerging as

a distinct ethnicity, a Gypsy-like or indeed `Gypsy' (according to local ascription) community (see

Marsh, 2000). The Ottoman Empire during the classical period also seemed to provide evidence that

the primary mechanism for the development of identities was through occupation, in the socio-

economic organisation of the Ottoman polity. The modern group identities of Sepetci, Kalayci,

Bohcaci, Demirci (Bakset-weavers, Tinsmiths, Pedlars and Blacksmiths) and others in Turkey and

elsewhere in the region, clearly drew upon this inheritance in the formation of 19th century sub-

ethnic identities, and pointed to the importance of socially constructed notions of bounded

communities. The `Dutch school' (see Acton, 2004: 98-116 for a comprehensive epistemological

critique, and Matras, 2005: 53-78 for a linguistic challenge), broadly argues Gypsies emerged as a

result of changing social and economic circumstances in 16th and 17th century western Europe, and

shifting intellectual perceptions in the 18th rather than the migrations of groups with `Oriental'

origins. This is essentially an argument regarding the social construction of Gypsy ethnicity from

groups of marginalised and criminalised peoples who became vagrants and `masterless' men and

women as a result of changing Poor Law regulations and shifts in concepts of the `deserving' and `undeserving poor', and as threats to a "well-ordered society" (Willems & Lucassen, 2003: 283-

304). At the time of beginning this study, the notions of a socially constructed identity in such terms

seemed promising and offered an explanatory paradigm that appeared to address some of the

contradictions and seeming confusion that surrounded the questions of Gypsy identity, and definition of the basic question of `who are the Gypsies? '

Yet the attempt of Willems and Lucassen to incorporate the situation of Gypsies in the Ottoman

Empire and pre-colonial India through a comparison of the mechanisms of exclusion in the presence

or absence of a "well-ordered" society, raises a point they seem reluctant to define; was Ottoman

society strong, centralised and well-ordered or, following Barkey (1994), was it de-centralised and

apparently weak, though actually effective at neutralising opposition through incorporation of dissident elements? That ambiguity demonstrated an inadequate understanding of Ottoman social

18


organisation (common to others in Romani Studies such as Zoltan Barany, 2002: 23-48,83-111).

Their description of Ottoman society as ordered "... differently from Western Europe... Instead of

local civic communities within a common judicial system, Ottoman villages and cities were

organised into millets with religion as the ordering principle... " clearly ignores the vast corpus of

civil law as promulgated by the sultäns, kanuni and the legal codes that were administered by the

kadis or religious judges, both of which were regarded as part of the complex common judicial

system applicable across the Empire, whatever local custom and practice additionally entailed

(Willems & Lucassen, 2003: 302). This undermined their previous position of a late 18`h century

establishment of Gypsy ethnic identity as a result of literary `fabrication' by stating "Gypsies were

the only category that was constituted on an ethnic basis" (also patently incorrect as even a cursory

reading of the Ottoman population records will show; 2003: 305). The chance to explore those ideas

and measure them against the more `traditionally ethnic' accounts has become a more central theme

of this book than originally intended, partly as a consequence of the debate that has continued in

Romani Studies on the questions of ethnicity and identity sparked by the Dutch scholars, and partly

as a result of the direct fieldwork undertaken in support of this inquiry, in Turkey and the region,

where these questions are perhaps more contested than anywhere else in the world. The initial

attractiveness of the social constructivist approach foundered in the face of both, and the realisation

that in an attempt to deconstruct the ethnic `grand narrative' of Roma ethnic identity, becomes its

own `grand narrative' of modernist and post-modernist identity construction. The application of

what might be described as ̀ Fish's razor' (the problem of making any present historical assertion on

the basis of arguing that all previous historical claims are radically contingent and vulnerable to

deconstructive analysis of the assumptions upon which they rest; Fisher, 1989: 304), highlights the

problem and I suggest that the Dutch social constructivists are the conceptual step-children of 19`h

century Gypsy Lorism.

In pursuing an investigation of Gypsies in the Ottoman Empire, I was concerned that the discussion

about the migration of Gypsies into western Europe has often relied upon a notion of flight from the

Ottoman Empire's rapid expansion as a matter of course (see the discussion in Fraser, where he

details the various accounts of the origins of the bands that arrived in western Europe in the early 15th century and only one mentions the Turks as responsible for driving out the Gypsies from their

lands, whereas the other accounts all suggest the Saracens i. e. Syro-Arabians, as the `push factor' in

this migration; Fraser, 1995: 60-78; Daniel, 1984: 28). The consideration of the historical context of

such migrations had initially led me to question whether the primary motive for the western

19


European arrival of the Gypsies was in fact the Ottoman advance (a process that had been taking

place for over fifty years in the Balkans and was almost certainly complete by the time of the

second battle of Kosovo Polje in 1389 CE), or whether there were other factors involved, connected

with the interregnum in the Empire following the defeat of Bayazit Yildirim ("lightning"; 1389 - 1402 CE) at Angora (modem Ankara) and the re-imposition of feudal Christian rule for a period

(1402 - 1444 CE) that almost exactly coincides with the arrival of the Gypsies, led by their counts

and dukes, in western Europe. The prospect that such an alternative perspective might offer opened

new avenues of inquiry related to the question of slavery and the relationship between Gypsies and

the feudal lords of the Balkan lands, especially as such a discourse feeds all too neatly into

predominantly nationalist conceptions of the period of eastern European kingdoms defending

themselves against the infidel (the ideology of the antemurale Christianatatis, the `bulwark of

Christendom'; see Norris, 1993: 258; Schäuble, 2006: 1-14; Hubel, 2006: 1-9), a common trope from

Serbia, Croatia, Hungary, Poland and the Vatican during the mediaeval period and revivified in the

conflicts of the Balkans during the 1990's.

Modern conceptions of Gypsy `trans-national' European minority identity are predominantly

articulated around a confessional adherence that is principally Christian and increasingly

evangelical (see Strand, 2001). In this sense, the notion that the majority of Gypsies were historically Muslim and not Christian both challenges the more Hindutva minded, and Pentecostal

identifying modern Roma identity. This is an uncomfortable position for some (at a Romani Studies

short course in Istanbul, 2006, one of the participants remarked that he understood what I had

presented in terms of the arguments for the importance of Islam in Gypsy history and that the

majority of Gypsies had been predominantly Muslim in the past, he could even accept the notion of

many of the traditional practices that I had argued were Islamic and not Indian in origin, but he

didn't like it). The concomitant idea of the historical flight of the Gypsies from the Muslim "terrible

Turk" is much more appealing to modem Romani (and European) politics than the notion of small

groups removing themselves in the chaos following Ottoman collapse, from the possibility of

enslavement by Christian princes with a desperate need for labour in their war-ravaged domains.

The discussion in this book began to move from one that I envisaged as attempting to investigate

and contemplate the context of such developments, within the cycles of events that we might describe as Ottoman history, to one where the questioning of the sources had become a broader

interrogation of the relationship between the variant narratives and history or histories.

Since these early departure points, the circumstances in which this study has been written have

20


forced me to reconsider almost every point that I began with. My relocation to the Turkish Republic

(and confrontation with Turkish nationalist, ethno-centrist historiography in contrast with Eurocentric historiography) and Istanbul in particular has meant that the entire perspective of this

book has shifted, literally eastwards. The many encounters with modem Gypsy populations in

Turkey and elsewhere in the region (Egypt, Bulgaria, Cyprus) has occasioned major alterations in

how I have conceived of what was once a rather dusty exercise in exploring the documentation

related to the late Ottoman Empire (tax and population registers, military cadastres and reform

edicts), to one where the actual stories of people's ancestors, the patterns still discernible in social

and economic organisation and the living heritage of these communities (to say nothing of the

actual diversity of these that remains a fact in Turkey) has profoundly altered the original plan. During the course of this work, not only geographical shifts have encouraged changes in

conception, but the experience of teaching Romani Studies courses in various places has born fruit

from discussions and debates with students, other scholars at conferences and seminars, and has

also had a profound impact upon how I have attempted to answer their questions of me. Research itself has been a primary motor for this change and in this respect, the opportunity to carry out extensive research in the field amongst the many and various Gypsy communities of Turkey and elsewhere in the region has frequently forced me to reconsider what I have `known' about the

subject, often resulting in new lines of inquiry and shifts in emphasis that would not have been

apparent otherwise.

The developing and dynamic exposure of the Turkish Gypsy groups to other European Gypsies and the European Roma awareness of the position and situation of Turkish Gypsies, has also played a major role in re-conceptualising the parameters of this study, as an exercise in viewing things from

the "other side of Europe" (the title of the 2"d International Romani Studies Conference in Istanbul, Bilgi University 13-15 May 2005). The notion of addressing the other half of the story of Gypsies in Europe, first suggested to me by Angus Fraser in 1998 at Professor Thomas Acton's inaugural lecture at Greenwich University and `pulling' attention and scholarship eastwards has been behind the majority of my work ever since. It was behind the motivation to co-organise two series of Romani Studies seminars and the first two academic Romani Studies conferences in Istanbul (the first at the Swedish Research Institute in Istanbul, 10-12 April 2003; see Marsh & Strand, 2006),

publish articles, research reports and an edited collection of works primarily dedicated to Gypsies in Turkey. The research projects that I have been engaged in over the previous six years have been focussed upon extending that knowledge and presenting it to the wider field of Romani Studies,

21


firstly as a doctoral scholar and more recently as a researcher for various projects funded by

institutions such as the Consulate General of Sweden in Istanbul (together with Elin Strand), the

Open Society Institute Assistance Foundation, British Council in Turkey, the UK's Economic and Social Research Council, and the European Roma Rights Centre (together with their local partners,

the Helsinki Citizens' Assembly and the Edirne Romani Association). All of these have made me

aware of much wider issues than those that are frequently afforded to historians engaged upon

research and the influence of these projects upon the developing positions that I consider, are

manifold. The extension of these parameters has also been a consequence of some re-considerations

of the fundamental questions of early Gypsy history and my attempts to answer the questions of how and why the Gypsies came to be.

These aspects are, I would continue to argue essential to understanding the arrival of groups of Gypsies in 15th century CE western European territory, and in seeing the development of what were

an èastern' people, forged in the complex milieu of I1`h century CE (4`h century AH) Sunni Islam,

Armenian Miaphysitism and Byzantine Orthodoxy. Attempts to locate the Gypsies in the cultural

and historical context of the `west' are fundamentally flawed by failing to acknowledge the importance of Muslim military and social organisation upon the earliest history of the Gypsy

peoples and in this context, the assertion that Gypsies are a European people is problematic (see Hancock, 2006: 70; Fraser's 1992 monograph is part of the "Peoples of Europe" series, though in

this sense the definition is more catholic since monographs about the Mongols, the Goths and the Armenians are also included in this series; Morgan, 1986; Heather, 1996, Redgate, 1998). Concern

to respond to the criticism of post-modernist scholarship that Romani Studies has merely observed socially isolated, marginal groups and exoticised them with an Orientalist identity, may lead

researchers to the position of too-readily refuting the èastern' origins of the Gypsies, without problematising the dualist essentialism of such definitions of èast' and `west'. This book attempts to both investigate these notions, and to `map' some of the parameters of discussion regarding a critical historiography for Romani Studies and outline what might be further questions for investigation into the processes of all identity formation.

In this sense, the present book is a very long way from where it began both conceptually and historically, and the considerations I have been forced to undertake have also meant a major shift in

thinking about the very validity of the undertaking, as I had originally conceived of it. I began as a historian looking to find data about a particular group in a certain time period through an investigation of the sources, an archival exercise preponderantly concerned with questioning the

22


texts. As such, this was perfectly within the `traditional' bounds of the discipline and certainly in

line with standard methodology in historiographical terms. Comments from my supervisor, Dr.

Thomas Acton that historians represented under-theorised sociologists were met with frank

rejection on my part, an insistence that history (or perhaps History) should stand apart and be seen

as an entirely legitimate discipline in its own right. Through the process of writing and researching

for this study and others that I have undertaken, I now recognise that Dr. Acton was right in this (as

in so much else), and that modem historiography has indeed sought to question the very

foundations that History is built upon, with profound implications. If historians are to practice their

craft (to paraphrase Marc Bloch: 1954), then the notions of history as objective, concerned with facts

and data and a patient `archaeology of the past' can no longer be held to be self-evident. The

challenges of post-modem inquiry and the post-structuralist considerations of language, discourse

and culture have to be met and some answers to the questions White (1989: 293-302) put forwards;

how do we as historians deal with the material we choose and select, the interpretation we place

upon it and the framework that we place around our very conceptualisations? In Romani Studies,

such considerations remain the untested ground of scholarship (though arguably part and parcel of

other historical enquiry some time ago). The impact of scholarship that seeks to address some of

these questions in the context of identity and ethnicity for example is frequently charged with its

relationship to the wider political considerations of Roma rights and emancipation or equality. This

relationship is problematic, as it both seems to empower those seeking to establish authenticity as a

means of securing rights in the context of nation-state minority policies, yet undermine the notion

of reliability as it appears to subvert scholarship to the requirements of a political agenda. These

considerations are something that I have tried to investigate in part of this book, when looking at

methodological and theoretical foundations, in the first section.

In some ways this study maintains a broadly `historicist' (in the Hegelian sense of the word), or

perhaps ̀ new historicist' structure (see Veeser, 1989: ix-xvi; White, 1989: 293-302; Fish, 1989: 303-

16) in attempting to `map' chronologically the periods of early Gypsy history through Ghazndvid,

Saldjük, Byzantine and Ottoman epochs, and seeks to present some suggestions as to the context of the development of Gypsy identity and ethnicity, in Chapters 3-7. However, I have tried to avoid the

use of an objective `third voice'; these discussions are presented by me as possibilities that may go

some way to illuminating the processes at work and would certainly fall into the category of "unwarranted speculations" (Schama, 1991). The entire book is in this sense a construction, what Jonathan Swift once described as something that contained "a great deal more plausibility than

23


truth" (Stott, 2007: 264), in that it argues explicitly that previous narratives are contingent and

therefore subject to deconstructive analysis of the assumptions that underpin them (and that such

assumptions have been suppressed or elided in order to maintain the illusion of veracity and

objectivity). I also argue that such an approach contains a tension, in that these previous narratives

rely upon exclusion and `forgetting' and thus can be seen as conscious misrepresentations, but that

such a position is implicitly undermining as it creates its own exclusions and treats all narrative as

suspect (Fish, 1989: 303-16). I have taken particular chronicles as my point of departure (mostly

texts relating to the Ghaznävids and Saldjüks from Islamic authors) and emplotted a narrative

around them that may, or may not present a different understanding of the circumstances

surrounding the processes of ethnic identity formation (to use the notion of emplottment from

White, 1975) from a perspective that seeks to draw attention to my own basic contention; namely that the early history of the Gypsies is intimately bound with Islam and the various power-states that it produced during the first one-thousand years of its history. Such a perspective underlies the very nature of this inquiry, and certainly underpins the notion that it is impossible to understand the

various manifestations of linguistic, social and cultural expressions that are associated with the Gypsy groups in the world without understanding this Islamic past. Such reverse teleology is again part of the broader and now questionable assumptions of history as ethnicity, and it is with profound misgivings that I am continuing the time-hallowed processes of seeking the present understanding of the ethnic community in a narrativised ethnic past. I have no solution to these problems, other than to relate these concerns and acknowledge that as a historian, history is what we choose it to be through our very interaction with what we investigate, to clearly outline here what those choices may have been and with what biases they were made and suggest that the metacritical questions that I pose about narrative may shift to normative constructions of the narrative itself in the process.

Beyond the more traditional approach of chronology and periodisation that I present, I have tried to raise the questions of theory and methodology in a more discursive and self-critical consideration of the study and the practice or research amongst the Gypsies in general, in particular the question of Romani history writing. I have placed this at the forefront of this study in order to attempt to unsettle what might be a far-too-complacent rendering of the history of the Gypsies, even if I have

sought to elaborate upon questions that seem to me at least, to have been little considered in the process of developing and constructing a plausible account (though one that is clearly not `true' in the sense that Borges or Swift suggest above). These considerations may serve to contextualise what is an ongoing process or series of processes of thought, and may be less than satisfactory in

24


providing new knowledge rather than presenting further questions. It is my intention that they serve to constructively critique my own position (and that of others) that offers apparent certainty. In this, I am profoundly influenced by the great American historian of European and Native American

encounters Calvin Luther Martin (1993: 3-4), who openly admits that "... words. I have grown increasingly suspicious of them... and am growing increasingly distrustful of what I myself have

been saying... " To attempt to use the discourse of western European scholarship and research to describe peoples and places, to bound together cultures and concepts that were very different,

encounters and understandings that were conceived of with other languages and another metaphysic is to impose a vastly different story than that told by those whose experiences I have purported to

convey here. The Gypsies in this story may well become "... the victims of illusions produced by

words" as Marcel Detienne suggests (1986: 26) and as a people, the Gypsies in all their variety and differentiation are the most victimised of all groups that have been fashioned by the words and illusions of others...

25


1. Historiography: History, Ethnicity & Identity

"The dust of many crumbled cities Settles over us like a forgetful sleep,

but we are older than those... "

Celal ad-Din Rumi (Mevlana)

HISTORY IS IDENTITY, the primary means of acknowledging sameness, membership of the group and

difference from others. It is always established, whether in part or wholly, through the sharing of a

narrative of origins, of journeys of migration, or anti-migratory narratives of autochthony, and of

`present' as related directly to `before'. Shared origins in the heart of northern Europe for a number

of peoples, such as the Swedes, Danes, Norwegians, English and Germans may be acknowledged

historically or archaeologically, but it is the trajectory of the narratives of journey from this point,

both geographically and conceptually, that begins the discourse of identities and ethnicities for these

peoples. There is no "geography of significance", as Maja Frykman (2005) has called it, in defining

this as a common point for the Swedes, the English or Norwegians; in fact these narratives of

journeys are frequently and positively anti-diasporic in their conception of origin. There is no

appeal to the "folk-wanderings" of proto Anglo-Saxons, Jutes, Svea or Göta as a building block of

common identity, no attempt to create a sense of commonality, whatever the linguistic connections

that patently indicate otherwise. Instead, notions such as the "cradle of Sweden" or dates (1066 CE

in England), are treated as axiomatic in the narrative of identity. Ethnicity, as a central component

of identity, is frequently established through what Siän Jones has called a process of "archaeology"

(1997; see also Sonia Tamar-Seeman 2006), the attempt to demonstrate the existence of direct lines

of inheritance from the present-day group to the past occupation of territory, and a common culture,

echoes of which are to be found in the artefacts and cultural expressions of the modern ethnos. Again, notions such as the "ancient Britons" - the Eceni, Brigantes, Trinovantes or other pre-Roman

groups, or the Cwen or Kväner (a minority group in northern Sweden and Norway, that frequently

adopt a counter-narrative to the Saami assertions of indigeneity in the Swedish case, arguably due to land-rights issues; for arguments concerning indigeneity see Onsman, 2004: 7-19), become integral

to interpretations of modern ethnicity and to a direct lineage with an ethnicised past (thus

effectively ethnicising all social, economic and political issues and actually undermining social

equality within the discourse of equality of opportunity).

Identity and ethnicity are then history, the narrative sum of the series of past events ascribed to

particular groups, and given legitimacy through the "major... democratic contest" of defining

culture, in what Edward Said (2004) described as "a disputed history of identity". In a way that

26


clearly transcends notions of identity and language as the fundamental nexus of identity, or a

common culture, shared religion or other criteria, history as it is constructed in narratives of origins

is the major conceptual framework for identity and ethnicity.

Is there a Gypsy history? A record of a whole series of past events associated particularly with

people defined by themselves as Romani, Dom, Lom, Travellers (or a plethora of other associated

terms), or more often by others, as Gypsies? The question may seem one that is self-evident to those

scholars and researchers working in the field of Romani Studies, but I would suggest here that it is a

necessary one, essential even, especially in the context of who produces this research and

scholarship, who it is produced for and why. A great many words are written purporting to describe,

define and delineate what is suggested are histories of the Gypsies (and thus establish legitimacy

through the past, or series of contested pasts), sometimes by Romani authors themselves (although

most frequently not), but as the poet David Morley writes, these are "... haunted by falsehood from

the start ... Fiction was the poached life-history of travelling folk" (Morley 2007). We might take

this as a leitmotif, as we concede that what is presented as research about Gypsy peoples, what has

been "poached" from them, in fact is more likely to be the record of contact between Gypsies and

non-Gypsy people with their imaginative re-construction, or fiction they define as "Gypsies... [with

their]... fantasies and longing for disorder" (see Malvinni, 2004: 23; van de Port, 1998; Okely,

1983). In this `history', we can find a record of racism, the mechanisms of misconception, prejudice

and exclusion and attempts to construct narratives of journey as an explanatory device for

discrimination (thus justifying the criminalising of mobility in sedentary nation-states and of course,

promoting the dominant trope of Gypsies as ̀ wanderers' or purposeless travellers), an exoticised

and orientalised version of groups of people who have actually been in proximity to others for

centuries, mostly through the experience of sedentarism. The idea of Gypsy identity being

confusing or indefinable is posited with very little comparison to any confusion of other identities

(the question of ethnic identity in any other group is rarely pursued in this context), yet we may trace `Egyptian' identity to Constantinople from the second half of the 11th century (see the many

repetitions of the story of the Atsinganoi at the court of Constantine Monomachus c. 1050, in

literature about Gypsies, originally from the Life of St. George the Athonite written in twelfth

century Byzantium; Peeters, 1922: 102-104), arguably earlier than the establishment of `English',

`Swedish' or many other European identities. The variation in origin myths that have abounded from quite early periods, have ascribed the most banal or bizarre of explanations to the ethnogenesis

of Gypsy people (see Mayall 2004; Hancock 2002; Fraser, 1992 for discussions of these). Words

27


then are not to be trusted, are fictions as Calvin Martin suggests (1993: 4). History and historical

research is "... a discourse... cultural, cultivated, fabricated and thus ultimately arbitrary... "

(Jenkins 1995: 12), a way of delineating the parameters of discussions about, in this case identity

and ethnicity, and ultimately access or restriction.

The notion of Gypsy history is one that is not secure though, academia has not always been

accepting of the legitimacy of such (much as other areas of study have been `ghettoised'). The

`establishment' in this instance might be defined as Historians, academic practitioners of writing

History, and in ways similar to those contests that have marked the definition of other "hidden"

groups in Sheila Rowbotham's seminal phrase (1972; see also Tebbutt 1998 for a recent example in

Romani Studies), Gypsy history has been frequently suggested as ̀ missing', `lost' or `forgotten' (in

the sense that the establishments of hierarchies of history "... forcefully excludes what it does not

embrace"; Terdiman, 1989: 227). The idea that Gypsies have little history has been extremely

influential, and is behind some of the misapprehension of non-Gypsy peoples about them. Ian

Hancock (2001: 10) notes what he describes as "... the vague understanding of Romani origins... "

and other writers have implied ambiguity (Fonseca 1995), or Gypsies as being without legitimacy,

through this lack of history (see Beck 1986, for his discussion of this as an aspect of perceptions by

non-Gypsies and researchers in Romania during the Soviet period). In contradistinction to other

histories, conceived of as the absent object of inquiry and signified by their remaining fragmentary

traces, organised (produced) by professional historians, archaeologists, archivists, librarians and

academics (according to Tony Bennett's analysis of the "historical past"; 1990: 22), the Gypsy `past'

is a lack of history behind, as Hancock (2001: 2-6) argues, the ability of non-Romani people to

ignore or leave out Gypsies from many aspects of society, "in the absence of a well-recognised history and clearly understood ethnic identity". Here we might suggest Brian Belton's phrase of a

people outside of the empire of written words, as apposite (Belton, 2005b; see below).

Historical research however, may be argued to be irrelevant to some Gypsies themselves in this

context. To know the family lineage, the relationships between groups and the status of those

relationships, whether cordial or antagonistic, might be what is important though frequently absent from the kind of research that concentrates upon resolving `problems' or `challenges' to social inclusion. To know whose family one's own ancestors once travelled with, or married into, these

things may have meanings; and the idea of an abstract record of the events stretching back into the

past, as a symbol of collective identity, seen to be of the non-Gypsy world (as Monica Kalderas of the Romska kulturcentret i Malmö [Romani Culture Centre of Malmö] told me on one occasion in

28


2004). This is the language of nationalism, of imagined homogeneous communities tied to

territories, of conceptions about when towns, farms, rivers, mountains and valleys stop being one's

country, to become "one's un-country" (le Guin, 1980: 211). In the perspective of national identities,

what is Gypsy history? Is it a pan-European or even pan-global history? The demand for

understanding the past of particular groups, through constructing narratives of ethnicity and identity

is part of a discourse of resilience and authority, of claims to resources or rights based in linguistic

conceptions charged with non-Gypsy notions of place and even time. The intellectual constructs of

many non-Gypsy scholars are those that are employed in an attempt to encompass experience and

events that are without the socio-cultural matrix of the academics and researchers producing

research reports about Gypsies (as an imaginary construct), for the most part. Those of us writing

history should be constantly mistrustful of what we say, what we describe as we seek to elaborate

the fragmentary glimpses of Romani people set down in non-Romani records, as we construct a

narrative of events that links movement with meanings, time and what has transpired. Our desire to

make a coherent picture of the past, one that we can refer to when faced with demands for

explanations as to who, where and how is, in its very inception, an acceptance of the legitimacy of

such logic, to implicitly acknowledge the notion that authenticity relies upon demonstrable

chronologies, maps and recorded `evidence'.

Yet, are we in danger of creating a new kind of essentialism, one that suggests that this process is

flawed and fraught because it has been produced by non-Gypsy people to non-Gypsy conceptions

and must be re-written by Romani scholars to be authentic and legitimate? The debate between

scholars in the recent past has clearly been contested over this ground and there are suggestions that

in the interests of the Romani emancipation movement and political activism associated with

securing rights for Romani people, this is the case (see Matras, 2002: 193 - 209). The notion of Romani history itself is an exclusivist approach, one that presupposes a unique Romani perspective that can be discerned from others at points in the past, elucidated from documentary evidence and textual sources. For an historically non-literate population for the most part, this is a position that is

clearly open to question. The perspectives of those who recorded the encounters with Romani

peoples historically are the dominant ones, even when they are directly quoting Gypsies themselves,

as in Andrew Boorde's 1547 The Fyrst Bake of the Introduction of Knowledge... or Lionardo di

Niccolo Frescobaldi's account of meeting Gypsies in the Morea, in 1384 (1818: 72-3). Gypsy historiography hasn't yet addressed the textual implications of the writing of that history, as it

simply relies upon a nomological or narratavistic approach (frequently both), whatever the

29


underlying ideological position of the authors. Scholarship about Gypsies is almost always

produced by `outsiders', non-Gypsies and many works have been instrumental in defining much

that we accept as the bedrock of Romani history and culture. Others have been significant in

defining what many regard as the propagation of stereotypes and anti-Gypsy prejudices (see

Arnstberg, 1998; Svensson, 1993) and their critics have challenged and highlighted this aspect of

their works (Montesino, 2002; Hazel, 2000; Strand, 2001: 195 - 199). To challenge racism and

discrimination, is it inevitable that an essentialism based upon equally exclusive notions of

belonging be created, or a counter-narrative of ethnicity and identity be constructed?

If the record of the past that exists is one that largely misrepresents this experience for Gypsy

people, is there a corpus of Romani historiography that addresses these misconceptions and

misconstructions? Increasingly the presentation of Romani histories is one that is being undertaken

by Gypsies themselves, and there is a body of work that we can define as Romani historiography

being added to those narratives of Gypsy people recorded and interpreted by non-Romani authors,

especially around key recent historical episodes such as the experience under Stalinism, or Nazi

atrocities against Roma and Sinti in occupied Europe (see Lewy's controversial 2000, The Nazi

Persecution of the Gypsies, where he argues that Gypsies were not targeted by the National

Socialist regime on the same basis as the Jews and Hancock's earlier counter argument from 1989,

Jewish Repsonses to the Porrajmos). The debate over the use of the term "porrajmos" or more frequently "0 Baro Porrajmos" - the Great Devouring - has meant that some activist Roma in

Sweden have adopted an alternative to the term "porrajmos", which they consider to be extremely

sexually offensive, using instead the phrase Sa o Mudarimös or Sa o Mudaripen "the final killing",

which they consider more accurate in conjunction with a reconsideration of the mechanisms of

exclusion, or what they term anti-Romaism. Hancock has cogently argued for the continuing use of

the former term upon linguistic bases; (2006a: 53-7).

What are the problems of a Romani historiography? Clearly, a consensus over terminology might be

one as suggested by the example above. What are the issues that confront researchers and scholars

writing Gypsy history or histories, at present? The notions that practitioners of history writing have

attempted to address, particularly as a result of the challenges from post-modernism, post-

colonialist theories and subaltern studies that challenge the legitimisation of a conceptual framework for professional historical enquiry itself, have been almost absent from many of the

recent works engaged with Romani history (Incirlioglu's critical engagement with Foucault's

concept of hetaerotopia, is an example of research that has attempted to address wider conceptual

30


issues, 2006: 191-201; see also Seeman 2006). Post-modernism itself has been challenged by some

historians (see Domanska, 1998: 173, when she suggests "I am grateful to post-modernism for many

things, especially for giving me an alternative apprehension of the world in terms of difference and

continuity rather than binary oppositions, but I am tired of ontological insecurity and

epistemological chaos. I need order. I miss metanarrative. ") As Thomas Acton has often remarked,

post-modernism is actually its own metanarrative premised by the acceptance of notions of pre-

modernism and a modernity that has been superseded (Acton, 2004: 103).

As David Mayall notes in his study of Gypsy identities, this may be less a lacuna on the part of

those writing Gypsy history, and more to do with the relative absence of historians in Romani

studies, until recently (2004: 26). It can be said Romani Studies scholars have frequently ignored

that the notion of history (Guy, 1975: 203). The works of Gypsy history that have been published

have often reflected a perspective that might be described as "alarmingly a-historical" (Mayall,

2004: 26) and at worst as an exercise in myth-making, yet the process of historical writing in general

is one that has encompassed much of the latter in the development of national histories (I am

reminded of Konrad Bercovici's statement that "... every historian has lied when telling the story of

his [sic. ] own people, and lied again when telling the story of another... ", 1928: 7). The shift away

from colonial and imperialist narratives of the 1911 century, to a post-colonial, critical historiography

that has been concerned with issues of representation - or its lack - and the subaltern `voice'

('history-from-below' in 1960's Britain, for example; see Past & Present, 1952- and History

Workshop Journal 1976-, for lively debates on the changing nature of history writing from Raphael

Samuel, E. J. Hobsbawm, Lawrence Stone), has been one that hasn't effectively answered the

questions raised by dramatic changes in political complexity in south eastern Europe, for example. As Milena Dragicevic Sesic (2005) has argued, the discourse of diversity in the region, fostered by

historical inquiry of the previous three decades, international organisations and others in an attempt to address the results of the conflicts of the 1990s, stands in direct opposition to ethnic-based

cultural policies and national cultures. The significant others in this context are the neighbouring Serbs, Bulgarians, Albanians, and Rumanians, but Gypsies are also part of the narratives of

ethnicity throughout the Balkans as van de Port (1998; see also Stewart 2002a: 185 - 190) has

shown, frequently as the ultimate `other' against which ethnic identity in any of these instances is

measured or `forged'.

There are then, a number of issues to be addressed by Romani researchers producing Gypsy historiography. Can these be seen differently for Romani writers of Gypsy history, as opposed to

31


non-Gypsy authors? Here I have tried to reflect upon aspects of the writing of Romani history as an

example of research and representation in a historiographical context, and argue the case that there

are a number of significant issues to be addressed by Romani researchers themselves. The first is

that Romani history is being produced differently by Romani and non-Romani authors and that it is

being defined through practice, whilst the debates concerned with questions of what history is

remain largely outside the purview of scholarship in Romani Studies. I am suggesting here that

there are apparently competing demands between Gypsy activism in the political sphere and the desire to construct coherent narratives of ethnicity, identity and history in the interests of addressing inequality, and the concerns of researchers attempting to examine the historical or contemporary experiences of Gypsy peoples. There is a relationship here, and the question is one of complement or conflict. I would suggest that the necessary engagement with the theoretical implications of modern historiography, for Romani history writing, are ones that must be undertaken, as part of the

shift towards a more critical Romani studies. This shift is one that the Romani writers of Gypsy history may be best placed to undertake, for reasons I shall discuss below.

32


1.2 The "Grand Narrative" in Romani historiography; from the "Great Trick" (o xanxanö barö), to

"We Are the Romani People" (Ame sam e Rromane dzene), and the Roma writing of Romani

history

"I am bound to tell what I am told, but not in every case to believe it... " Herodotus 'Histories'(] 862)

FROM ANGUS M. FRASER'S SEMINAL 1992 HISTORY OF THE GYPSIES, to Ian Hancock's 2002

monograph, lies not only a temporal separation, but a paradigmatic shift in the approach to the

writing of Romani history. The metanarrative or "grand narrative" (to use Lyotard's term, 1979:

xxiv-xxv), of Fraser's chronological progression has been refashioned by a differing conception of

the structure ("0 Teljaripe: The Move Out of India"; "0 Aresipe: Arrival in Byzantium & the

West"; "The Buxljaripe: Out Into Europe"; Hancock, 2002: vii - viii), with a positive emphasis on

the perspective of Gypsy people themselves, by a Romani author. Fraser's broadly thematic

approach ("Pressure of the Gyves"; "The Approach to Avernus", 1992: v-vi) describes a more

`traditional' historicism and maintains the fiction of the objective voice, reflecting something of his

attitudes as regards aspects of various notions of origins for example, and his characterisation of the

movement of Gypsies into central and western Europe as ̀ The Great Trick' (o xonxan6 barb as

Fraser renders it, though this is actually incorrect -o xaxaimos barb should be the phrase, as

Hancock suggests; pc. 2008. Acton has indicated that this is derived from Borrow, though Fraser

does not acknowledge this; pc. 2008). This, Fraser suggests, was "the greatest trick of all... played

on western Europe in the early fifteenth century" (1995: 62). This strikes a chord that is echoed

elsewhere in the work, one of stereotypical roguish mendacity when it comes to claims made by

Gypsy people about identity or belonging, throughout the whole of the recorded history of the

Gypsies. Though this is not the place for a review of the volume, the underlying trope is one of ironic scepticism, and the text is emplotted as a heroic but tragic metanarrative, clearly defining the

resilience and resistance of Gypsy peoples in the face of almost overwhelming oppression and

suppression. The link with resilience and therefore authenticity is, however, slightly contradictory in

view of the sceptical position adopted in Fraser's analysis. Hancock's approach is a post-modem

narrative by contrast, where the author stands firmly in the text in a dialogue with the reader (see

the Introduction, where he writes that the book is a way "... you can get to know more about us and

our culture, history... and aspirations", 2002: xvii). The connection with resilience and authenticity

remains, but the text is intentionally didactic and makes no attempt at preserving a fictional third

voice - the text demands that the reader engage and align himself or herself with it. Elsewhere

Hancock presents the reader with his own intellectual processes and developments in his thinking,

33


"... in my earliest writing I supported a fifth-century exodus from India, and accepted the three-way

Rom-Dom-Lom split; I no longer do" (2006: 69), in contrast to Fraser who dispassionately suggests

about the debate over origins for example, "so there we have some interesting coincidences"

(1992: 31).

The challenge of the post-modernist Dutch school (Willems, 1993; Willems, Lucassen & Cottaar,

1998, for examples), has forced a recapitulation of the arguments about origins and identity that, to

some extent were left open by Fraser's scepticism (see Acton 2006: 11). Ian Hancock's recent work

has sought to define the question emphatically, and has brought strong reactions from the social

historian David Mayall (2004: 223-224), and more particularly, Yaron Matras (2004: 53-78). In his

review article (2004: 199-209) he challenges Ian Hancock's claims to present a convincing case, and

argues instead that activism, rather than scholarship, is the driving agenda in this recent discussion

of origins and identity. Thomas Acton's response in Roma (forthcoming), suggests that the

arguments are in the nature of "classic positivist" debates, familiar in Romani Studies since the

arguments of W. R. Rishi (1976), began to represent the modern kshatriya (Rajpüt warrior) position.

The notion of a military origin of the Gypsies is of course nothing new; Sir Richard Francis Burton

KCMG, writing on a number of occasions (see Burton, 1898: 215-7, for the most definitive

presentation of his arguments) suggested this from the 1840's onwards (as Hancock has himself

made clear in a recent posting to the Roma Virtual Network in July 2007) and others followed for

the next sixty years to repeat or develop this. W. R. Rishi's own discussions sought to draw new inferences from some surprising connections between Panjabis and buffalo milk for example (1975). The writing of Romani history remains a contextual, and highly contested arena, where the

discourse of "authenticity" and "resilience" jostles with that of "social isolates" and "marginality".

Scholarship and activism are contrasted as two opposing poles, with the engagement in one argued by many as compromising the other (see Mayall 2004: 5-8; and Matras 2004: 62-78; 2002: 206). In

this sense, Romani `self-writing' (to use a phrase from a 2005 symposium at Umea University's

Department of Modern Languages), can be seen as the necessary corrective to gadje (non Gypsy)

derived scientific criteria, and positivist notions of objectivity. In this context, Gypsy researchers'

positions are very similar to that of other writers from minority ethnic backgrounds; it is the

assumption that the activist agenda is always to be identified at the heart of the argument, the inability to stand ̀ objectively' above the debate. The problem that such a position also embodies (in

that any attempt to pursue objectivity is seen largely as the product of an exterior, or

unrepresentative perspective that cannot adequately supply us, the readers, with an insight from a

34


genuinely ethnic voice), is of course the flip-side of this particular counterfeit - in the sense that

notions of scientific objectivity have been undermined by the successful critique of post-modernism

- coin.

In the wider Romani political movement, the `traditional' approach to the history of the Gypsies has

largely maintained its teleological narrative, through the tropes of journey, persecution, and the need

for redemption through political and social emancipation, delivered by non-Gypsy institutions

(European Union, Council of Europe, Organisation for Security & Co-operation in Europe and the

Open Society Institute), as mobilised by Romani activism, and influenced by research reports and

academic studies. Web-sites dedicated to the dissemination of information about the Roma of

Europe, frequently include varieties of historical background that continue to reflect the emphasis

on mobility and marginalisation (see the web-site of Patrin as an example). Contemporary music

and dance, as an aspect of the reproduction of what one might term popular Romani history,

constantly refers to the "Roads of the Roma", or the "Thousand Year Journey" (see for example,

Eriksen & Dreisziger, 2004), reinforcing the separate nature of Romani experience through

alternative narratives of journey, or counter-structures of community governance and self-

regulation, with collections of music by the "Gypsy Queens" or "Kings". Conversely, these seek to

integrate this cultural expression under the general category of "world music", again as an

alternative to `mainstream' (non-world? ) music. The possibility of Romani music that expresses

more "conventional" forms, such as the Mozart, Brahms and Liszt played at concerts by Robbi

Lakatos or Gabor Boros' ensemble, finds only a limited market, whilst those musicians who work in

a more nationally-defined genre, such as Swedish dance-band music, are not recognised as Romani

or Gypsy at all (despite the fact that very many of the dance bands' personnel are Resande, or Travellers). The extreme example of this is in the situation of English Romanichals, whose musical heritage has become almost wholly absorbed as "folk music" since the latter was re-fashioned in the

1960's, and English Gypsies now often identify closely with American country and western music. This form of Romani presentation as an expression of historical experience has become detached,

de-contextualised to the extent of being unrecognised as such in the Swedish and English contexts. The particular descriptions of the shifting relationship of Travelling peoples from rural to urban

communities, as a consequence of industrialisation and urbanisation, are now taken to be part of the

overall narrative of population dislocation and (frequently) emiseration, in the restructuring of these

nation-states during this process. In these examples, the research undertaken into these musical forms leaves aside such considerations (Scottish research by contrast, under the auspices of

35


Edinburgh University's Centre for Scottish Studies, has an unrivalled archive of recordings from

Gypsy-Travellers singers such as Belle Stuart and many others). Sonia Seeman has argued that

contemporary Gypsy music in Turkey, produced by the Romanlar themselves, is both responsive to

and reflective of the non-Roman "iconic stereotypes" that emerge in the struggle for control over

representations of the `Gypsy' (cingene in what is considered to be a pejorative term), in what she

suggests is a fluid "... contingent, negotiable and contestable ... relational and conjectural rather than

essential" process of establishing Gypsy cultural identities (see Seeman, 2006: also Marsh and

Strand, 2005: 12 - 16).

In the purveying of popular ideas of the Romani past, the imaginary Gypsy, and his/her connection

to "the wild" or exotic, maintains its hold on both the European conception of Gypsy people, and

the understanding of how they came to be. The current prejudice and discrimination displayed

towards Gypsies in Europe utilises this discourse in order to mobilise the notions associated with it,

underpinning stereotypical representation through the media of feckless, irresponsible parasites. The

portrayal of Gypsies in historical terms is significantly undercut by reference to de-legitimised

"travellers". A great deal of scholarship and research reports that are published engage critically

with this discourse, arguing for a rights-based approach that ultimately challenges some of the

conceptions surrounding notions of social inclusion and citizenship, especially as these are frequently ethnicised across social, economic and political factors further marginalising Gypsy

communities. The representation of Gypsies, Roma and Travellers in scholarship is as many-faceted

as the research itself, but the current plethora of texts that focus upon social inclusion is less an

indication of the needs of Gypsy peoples themselves, and more a reflection of where it is possible to

gain funding for scholars and what kind of research is being commissioned. In this context, the

production of scholarship is always an aspect of those producing it, not those about whom it is

produced. These many representations have a clear impact upon how people perceive Gypsies, and how Gypsies perceive themselves more and more. As Acton has suggested, the solidity of a rights- based model of research (such as that pursued by the European Roma Rights Centre) stands in stark

contrast to the smoke-and-mirrors that are preferred by many researchers and scholars, the

essentially nationalist notions of social inclusion and citizenship (2004: 114).

The epigram from Mevlana Celaladdin Rumi at the head of this chapter leads me to the consideration of Romani identity in methodology and research, and suggests one of the principle

phenomena that appear in these, namely the construction of Gypsies in the imagination of the

observer. Like beauty, the image of the various peoples described as Gypsies is frequently to be

36


found in the eyes of the beholder. The repetition of various attributes of Rom, Dom or Lom (or the

various ethnonyms by which groups are described as Travellers) that are the subject of research

(Lom and Dom appear all too infrequently in what is a predominantly Eurocentric discipline; for

examples, see Hanna, 1982; Vardan Voskanian, 2003: 169-80), concerning all manner of behaviour

that is defined in ethnic or cultural terms (both of which are often thinly disguised alternatives to the

less acceptable term race), is another. To the extent that these represent consistent methodological

concepts applied or implicit in research and scholarship, they are the dust of Mevlana's "crumbled

cities", or as Italo Calvino might suggest, the "imaginary cities" that academia has in the past, and

continues in the present in some cases, to construct as the sites for their interpretations of who and

what are Gypsies (1978). Within these carefully (or not so carefully) built edifices, Gypsies are

positioned, assigned the role of players in the drama of symbolic action that constitutes the attempt

to portray an understanding of their lives. This is `smoothed' over almost inevitably to ensure a

degree of consistency (and those of us engaged in field research will know that contradictions are

worth a great deal more than smooth consistency) and provide what are in the end, a series of

conclusions often designed to demonstrate the necessity for conclusions, interventions or strategies

almost always defined by a wider socio-cultural and institutional context, what Acton has recently

termed the "[... ] shifting and uncertain sands of `citizenship' and ̀ social inclusion' policies" (Acton

& Marsh, 2008; see also Acton's critique of the general approach and in particular, the UNDP 2002

report Avoiding the Dependency Trap, Acton, 2006a: 30-33). In more concrete terms, research often

seeks to identify Gypsies according to an a priori set of criteria - frequently based upon previous

academic research - before proceeding to observe the group, interview and collect data and

subsequently establish a framework in which to construct narratives of ethnicity and identity with

the prerequisite elements of music, dance, language, religion, and cultural practices.

Other aspects often make their appearance in research, such as the notions that Gypsies represent a

social `problem' (or more euphemistically, a series of `challenges') in terms of integration and what

might be described as the citizenship or social inclusion paradigm (see the numerous reports

considering this conception of the issues from largely uncritical [of the conceptions of social inclusion and citizenship] perspectives), such as this objective from the Government of the Principality of Asturias: "To approach real situation of gypsy [sic. ] community, its needs and their deficiencies... in order to jointly define proposals for the social incorporation of gypsies and to

move from the social exclusion to the real citizenship... " (Government of the Principality of Asturias, 2006). The questions that underlie this kind of research are about how well or poorly

37


Gypsies ̀ fit' into non-Gypsy societies (with the concomitant notion that there is a homogeneous

society against which they are measured and evaluated; see the work of Swedish ethnologist

Svensson, 1993; its critique in Montesino, 2002, and its antibook in Hazel, 2000). There is also a

clear concern about control of movement and migration, and in much research the question of crime

and its relationship to Gypsy communities is at the heart of the inquiry, frequently viewed through

the prism of statistics and quantitative data, or the distinctive gaze of the state (see ethnologist

Arnstberg, 1998, and the critique by Strand, 2001: 195 - 199).

The particular consideration here is how we as researchers, both Gypsy/Traveller and non-

Gypsy/Traveller, choose to portray (or sometimes betray) the communities and individuals we are

working amongst and how, in the context of academic research, we as a community of scholars and

researchers confirm, create or refute the prejudices, stereotypes and misconceptions that exist about

Gypsies, through our [mis] representations. In the sense that research, as I suggested above has a

very direct bearing upon knowledge and understanding, and ultimately policies and practices (and

more important, government spending, NGO budgets and philanthropists' donations), the

responsibility of researchers is one that is often treated lightly, though mistakenly so. Representation through research is the primary means by which other researchers and academics, international and national policy-makers, advocates and activists perceive the peoples we describe

as Gypsies, as if they were in fact real, and not the interpretation of the researchers who depict them

in the pages of theses and reports, in many cases debating over the representations as if they were in

and of themselves, a totality. These representations take the place of the actual people and come to

stand for them as symbols or sometimes ciphers for a series of notions (often graphically reinforced

through accompanying photographs, reifying individuals even further "... in picturesque discussion

with Roma and Dom in tea-houses, tents and elsewhere", in what Acton has recently remarked as "... a diversionary post-imperialist indulgence. " Acton and Marsh, 2007), related to the overall trope (metaphor, synecdoche, metonymy or irony, of which the dominant tropes in Romani research tends

towards synecdoche and metonymy) and emplotment (a verbal structure in the form of a narrative

prose discourse) being used by the writers - the frameworks by which the research has been

formulated (White, 1978).

What are the assumptions present in our research that we use to emplot our narratives, and the tropes that we construct? Clearly the Gypsy as metonymy in the social sciences imagination of much research is the major trope - in the case of the work of van de Port for example (van de Port, 1998: 6-7), Gypsies are understood to stand for `wildness', `licence' and a people `unbounded' by

38


what are perceived to be the conventions surrounding behaviour in the `majority' society, in this

case the Serbs of Novi Pazar. Van de Port is, in this instance, using the trope of Gypsy as a means of

examining Serbian people, emplotting his work through a narrative of the tragi-comic and post-

modern satire. In many cases, the use of the imaginary Gypsy is a device to examine the non-Gypsy,

to actually explore the psyche of the gadje. The 2002 UNDP report so deftly critiqued by Acton

referred to above tells us more of the conception of police officers, social workers, local

government administrators and UNDP researchers as to who they perceive as Gypsies and what

they understand or actually assume to be the criteria for defining them, than it does about Gypsies

as individuals and communities experiencing the particular circumstances in which they live. This is

a recurring problem with much of the research conducted by albeit (mostly) well meaning

individuals in the field of social sciences. It tells us about the people conducting the research, those

funding it and the audience it is intended for through re-presenting the Gypsy using tropes we have

come to expect - excessively poor, often itinerant, ignorant and under-educated, disenfranchised

politically and marginalised economically, socially excluded and culturally appreciated within a

very narrow context. Research that offers other perspectives is far less prevalent though of course it

exists; in a presentation at Istanbul Bilgi University in 2004, Elena Marushiakova and Veselin

Popov provided examples of a counter-narrative from their recent (at that time) field work in the

Crimean region and Ukraine that challenged the expectations of the audience significantly. Does

research that brings us information about those Gypsies who may present alternative or contrasting

perspectives to these dominant tropes have an impact upon the wider body of scholarship? Despite

the fact that such research does go on (Kinga D. Toth's 2001 doctoral research at Manchester

University about successful Romanichals in the UK, or Nidhi Trehan's 2008 doctoral exploration of

the notion of international non-governmental organisation personnel as elite groups and their

relationship to the Roma and Gypsy grassroots movements in Europe, are two examples I can think

of), little of this finds a resonance within wider social policy research apparently. So it would seem

to be the case that research, funded by academic scholarships or major NGO's, trans-national bodies

or national governments is concerned with presenting Roma, Gypsies and Travellers as Gypsies -a set of notions surrounding the researchers' ideas of who these people are. As Brian Belton has

remarked about research in a Czech radio programme, "It's an empire of written words. It's an

empire of writing that exists separately from people... "(Belton, 2005).

However, the notion that all research is negative and pervaded by stereotyped representations of Roma, Gypsies and Travellers is itself something of a stereotype. Stereotypes are specific products

39


of time and place that appear to offer simplistic and all-too-frequently negative, explanations for

specific phenomena, by generalising them and distorting them. Frequently following the word

"but... " stereotypes are results of these phenomena. Discriminatory and prejudicial stereotyping

about Gypsies is couched in these pseudo-explanatory terms, when in fact it is a product of

exclusion and marginalisation, not an explanation of them. Obvious as this seems, the frequency

with which these appear in both research and the responses to it is surprisingly high, and many

critics of researchers or their research, perceive both to be inherently stereotypical in their portrayal

of Gypsies, Roma and Travellers. As a recent conversation about a particular researcher amongst the

Dom of south eastern Turkey suggested, the stereotype of the unscrupulous, invasive and self-

serving scholar is widespread enough to reach even these largely unappreciated people. The explicit

comparison being made was with the research I was carrying out as a Romani person (though as a

thoughtful reflection on this mission made by one of the researchers with me pointed out, despite a

critical perspective regarding identity politics and the mobilisation of ethnicity as the prime factor

in resilience and the continuing resistance of Gypsy communities to the demolition of their homes,

marginalisation from education, employment and health services and a score of other problems, my

discourse during research was entirely bounded by the wider one of Gypsy ethno-politics in

general). The stereotype of the `bad' researcher is one that may be reflective of the very real

negative experiences of people on the outside of the "empire of written words", but it nevertheless

offers little in the way of explanation about why these should be so.

Research and researchers operate within significant constrictions during research projects or field

missions, and these may be some of the reasons why they seem to reflect the stereotype of `bad'

researchers. The determinants of the research are frequently in the hands of the funding bodies of

the programme, and as such this can have a negative, sometimes deleterious impact on the results (and thus the perceptions presented in findings and recommendations). Research criteria is often

telescoped to view a very narrow set of questions, ignoring the wider considerations that might

modify or radically alter the outcomes, and the question of funding is almost always a major criteria for how much, how long, how extensive or how frequently people are interviewed or who is being

interviewed (The UNDP report Avoiding the Dependency Trap illustrates the problems with this

question, whilst the 2004 report "The Situation of the Roma in an Enlarged European Union",

written by Focus Consulting Ltd, the European Roma Rights Centre and the European Roma

Information Office, demonstrates that these can be overcome with careful research design and

methodology; European Commission, 2004). In short, the funding and resources underlying any

40


research will often determine the kind of limitations that researchers operate within, and affect the

results.

This, of course, is not the only reason that researchers may present ambiguous or negative

perceptions of Gypsies through their reports or publications; in some cases the agendas of

researchers are apparent from the very outset in the titles for their publications (the example of the

UK's Communities & Local Government Office 2007 report, "Gypsy and Traveller Task Group on

Site Provision and Enforcement: Interim Report to Ministers", Communities & Local Government

where the section entitled "Community leadership" states, "We have identified skill and people

shortages in planning, enforcement and inspectorate agencies", clearly suggesting the importance of

ènforcement' with regard to Gypsies and Travellers). In other cases, the hidden assumptions behind

the research that surface in the kinds of questions asked, the kind of material gathered and the

conclusions drawn from it, tell us much about the views of those who conducted the research in the

first place and their intended audience.

The impact of historical contingency in many studies is frequently subsumed under a generalised

abstraction that is ultimately self-referential - Gypsies are a marginalised and socially excluded

group because they have always been so, because they are from elsewhere and òther'. The

relationship between modernity, the nation-state and exclusion is rarely examined as a context for

much research, possibly reflecting the perspective of the audience for many of the reports produced

(the work of Ussama Makdisi about sectarianism and violence in Lebanon is a notable exception;

1996: 23-26,30). The concerns of the modem nation-state and trans-national, supra-states are to

improve upon the model and ensure social inclusivity, rather than reflect upon the historical realities

of nation-state construction as inherently exclusivist in the promulgation of `the nation' and ethno-

nationalist ideology (for a reflective discussion of Roma `nation-building' and its consequences, see

Horvath, 2006: 53-57; for a discussion about the International Roma Union's Prague 2000 adoption

of the "Declaration of a Nation Without a Territory", see Acton and Klimova, 2001: 157-219).

The problem of reliance upon older analytical approaches drawn from linguistics, ethnography,

anthropology and social sciences being applied to historical phenomena, has been one that has

concerned me in the research for this book. It is one of the developments that has been part of the

process of researching and writing a text that might be best described by the term ìnter-

disciplinary' research, and has become the model for this particular study and those related to it in

the course of my work. At this point I would suggest that a more reflective scholarship founded

upon a critical, multi-disciplinary approach can be realised, otherwise Romani Studies is likely to

41


remain underpinned by late 19th century Euro-centric conventions of linguistics, folklore and `racial

science' (Fraser, 1997: 172), and an approach that exoticises and objectifies. It is deeply

unsatisfactory when a cursory reading of the most general works, the repetition of a ̀ standard'

discussion of Romani origins and the recycling of much that was first produced in the 18th and 19`h

centuries, can establish èxperts' and interpreters of Gypsies, of (usually) non-Romany background

(the encyclopaedic works cited elsewhere in this book are examples of this). The growing disquiet

about the increasing numbers of such ̀ experts' from both Romany and non-Romany people has led

to an identification of the so-called "Roma industry". Despite the impassioned plea of a Romanichal

activist and poet at a Greenwich conference of international scholars to recognise this (Frankham,

2000), the phenomena has increased, almost in inverse proportion to the general worsening of

conditions for Romani peoples in Europe (Harding 2001: xi-xii). Forced evictions, declining levels

of access to health care, education, employment and training are rapidly becoming the long term

`norm' for many of the Gypsy communities in eastern and south eastern Europe. Gypsies have

become the largest net export of nations in eastern Europe since the collapse of Stalinist

Communism (Stewart, 2002b), and attitudes in receiving countries in western Europe and the

United States continue to harden, as both popular opinion and political legislation are mobilised in a

discourse about illegal immigrants and refugees. Gypsy is, in this context an additional series of ill-

defined concepts about identity, equating with òtherness', àlien-ness' and `difference'. Such

notions are only reinforced by a discourse of Rroma nationalism that seeks to locate Gypsies as a

`non-territorial nation' rather than related minorities located in nation-states (according to the

International Romani Union's World Romani Congress Prague gathering, that approved the notion;

Acton & Klimova, 2001). Funding for social programmes aimed at addressing or alleviating the

myriad problems facing Roma and Romani communities in Europe have left few lasting benefits in

the decade and a half of post-Soviet politics, whilst attempts to constructively address the social

inclusion of Gypsies remain at the regional or national policy level, but seem largely ineffective (the

debate about this is at the heart of the `Decade of Roma Inclusion' initiative of the Open Society

Institute and certain East European governments). The choice some communities make to continue

to be òutsiders' is, as it always has been a response to rejection and thus no choice at all.

Not all research is restricted to the presentation of negative stereotypes or notions that reinforce

common prejudices of course; in some cases academics and researchers are keen to present certain data in the context of their own agendas for mobilisation or organisation of Gypsies, or to support

the arguments made regarding origins, ethnicity and identity. The research into Gypsy history is one

42


such area where competing agendas and conflicting interpretations may reflect this to a much higher

degree than in some others perhaps. In the sense that the audience of this kind of research may be

broader (in that much social science research is directed at a non-Gypsy audience and intended to

achieve change through mobilising it), it is an arena where a series of counter-narratives to the

dominant themes of the nation-state, social inclusion and citizenship, have been presented as the

historical experience of Gypsies. This representation of history has been made through the

construction of an alternative narrative of journey, of exodus even, from the Promised Land or Baro

Than, as Rishi suggests (1989: 1). The notion of the non-territorial nation is consequent with this as

it suggests a displaced people, exilic and wandering like the Hebrews in the Sinai, and the potential

for a return perhaps to a place where Gypsies are not alien and other. The question might be to what

extent does scholarship that seeks to define this original departure echo these notions? Has the

investigation of origins served the interests of Roma and other Gypsies in the confirmation of an

otherness that justifies the discourse of social inclusion and minority rights, through asserting that

like other minorities, Gypsies stem from elsewhere and are therefore inherently alien to the majority

of society, distant and removed?

In this sense, the experience of the Gypsies in the Ottoman and Turkish lands offers an alternative

perspective to history, as the emphasis (in modern Turkey) is on an identification as part of the

nation-state, not as a minority with a diasporic heritage. The Ottoman picture is harder to discern as

the records of how Gypsies in the Empire perceived themselves, are limited, though the occasional

glimpse can be found in some official documents, but these may have portrayed a particular series

of notions for a purpose (see Ginio, 2004: 117-8). They do however indicate a sense of identifying

with the broader society of Muslims, though clearly this is complex as it involves an oppositional duality with being Gypsy (Ginio, 2004: 118). In order to move on to the question of Ottoman

Gypsies and their particular place in society, I review some of the literature that has emerged

regarding Ottoman Studies and the nature of minorities in the Empire, and relate this to the work that has looked at Gypsies in Turkey.

43


1.3 Stout hearts and sturdy footwear; the Gypsy Lore Society and other literature

"... and who shall say what proportion of fact, past, present, or to come, may lie in the imagination? " H. Rider

Haggard, 'She'(1886)

IN THIS SECTION, I wish to examine the some of the developments in Romani studies and Ottoman

historiography that have a bearing upon this book, a brief historiography of the writing of Romani

history as it more particularly applies to Turkey. In critically assessing the state of Romani Studies

in this area, and in light of recent publications providing an analysis of the Ottoman Empire's

Gypsy populations, I shall discuss the concept of late Ottoman ethnic, confessional and social

organisation. This has often been referred to as the millet system (Bosworth, 1982; Braude,

1982: 69-88; Braude & Lewis, 1982: 1-34; Cohen, 1982; Davison, 1990: 112-132; Faroqhi, 1995;

Haldon, 1992). Some inaccuracies that have arisen within Romani historiography regarding this

subject are due to a lack of clear understanding about the Ottoman Empire in general, and the millet

system in particular (Braude, 1982: 70). There has also been little historical analysis applied to

Gypsies in the Empire beyond the European provinces (Marushikova and Popov, 2001; Ginio,

2004; Fraser, 1992: 171-176). Emphasis upon the Roma as a European people makes reference to

their `Indian origins', but frequently with little historical contextualisation, relying upon generalised

descriptions such as "ancestors of the Roma left their mother-country India... about a 1,000 years

ago during the Muslim invasions... " (Rishi, 1989: 1) and vague notions of the history of 11th century

Asia and Anatolia. It often appears as if little had been written that does not echo or uncritically

repeat many of the earlier texts from Romani historiography in the Journal of the Gypsy Lore

Society. This book is additionally written to try to begin to address some of these issues from a

historical basis in Turkey in particular. Relatively little has been committed to paper regarding the

Gypsies of Turkey, in spite of the importance of these lands in the history of the Romani people.

What has, has largely examined Turkish Gypsies as homogeneous except by rare references to

confession, language and more frequently to occupation. No particular studies have considered the

position of Gypsy women or children in and of themselves, and there is a tendency to consider Turkish Gypsies (as in a great deal of research in this field) as an undifferentiated mass, without

stratification in class terms, gender or age. The references to Gypsies in the Byzantine period will be more fully discussed in the following section, but in modern scholarship at least, the description

of these groups has been sparse. Here, a brief survey of some of the research concerning them is

undertaken.

Leaving aside for the moment the travellers' accounts that mention Gypsies in the Ottoman Empire

44


(such as Evliya celebi's descriptions from his Seydhatnämesi, of which there is currently no entire

translation in English; see Palles 1951; Dankoff and Elsie 2000; Dankoff, 1990), the first serious

attempt to analyse aspects of the Gypsy communities in Turkey come from Dr. Alexander C. Paspati

M. D. (also Alexandros G. Paspates) in the 1860's who attempted to describe the language in use

amongst them in his "Memoir on the language of the Gypsies as now used in the Turkish Empire"

in the Journal of the American Oriental Society, (1860-1863: 143 - 270). In his introduction to the

history of the Gypsies, Paspati refers to the fact that "no general persecutions ever took place

against them, either on religious or political grounds... " and as a consequence "they have been

suffered quietly to live in those provinces [of the Ottoman Empire]... and have multiplied to such a

degree that they are superior in number to their fellow-countrymen in all other states in Europe... "

(1860-63: 147; Fraser and others follow this closely; 1995: 175). He goes on to stress what many

European commentators also suggested in arriving at an estimate of the population, namely how

difficult this was as the Ottomans counted only sedentary Gypsies and usually undercounted

women, boys under 15 years and men over 60 (see Karpat, 1985: xiv); their numbers for Muslim

Gypsies are also unreliable. One total that Papsati derived from other, states that "three fourths of

all the Gypsies of Europe are to be found in Turkey... " although in actual terms, he is equivocal

about the figures derived from Gypsies themselves "who by such mendacious accounts are inclined

to give themselves importance and consideration in these provinces" (Paspates, 1860-3: 147).

Paspati considers this data important because it confirms that great numbers do indeed live in

Turkey, as referred to by travellers' accounts. He goes on to note that the Gypsies of Turkey follow

the religion of those whom they live amongst, and that they inter-marry with Turks but not with Christians (1860-3: 148). In the following pages Paspati proceeds to analyse the language of the

Gypsies after making his famous remark "The entire history of this race [sic. ] is in its idiom... "

(1860-3: 149), a maxim that might be said to have guided Romani Studies ever since.

In his opus, Etudes sur les Tchingianes ou Bohemiens de 1'Empire Ottoman, Paspati again refers to

population in Turkey and how numerous Gypsies are (1870: 3 5), before suggesting, in a rather

curious remark that the differences in dialects spoken in the Empire meant that Balkan Gypsies may

not understand those in other parts (1870: 36). This I would suggest is an indication that Paspati was aware of the fact that both Domart and Lomavren were being spoken by groups of Gypsies in

Turkey, but not the knowledge of what these ̀ dialects' actually were - this would follow upon the

work of Franz N. Finck in 1907 (34 - 60), K. P. Patkanoff in 1908 (229 - 257; 247 - 266; 1909: 325

- 344), G. F. Black in 1913 (327 - 330) and John Sampson's 1907, "Gypsy Language and

45


Origin"(422), amongst others that would effectively chart the varieties of "Asiatic Romani'. Such a

possibility is useful in that it may give us an earlier indication of the presence of Lom in Turkey

than the modem communities residing in the Black Sea region provide; these suggest an origin that

is to be found in the Russian ethnic cleansing of Muslim peoples from the Caucasus in the 1877-78

Ottoman-Russian war, in a discourse that is reminiscent of the mübadile (the population exchanges

of the 1920's and 1930's) narrative of the Romanlar population (from interviews conducted in this

region, September 2007). Paspati also draws a distinct linguistic difference between Christian and

Muslim Gypsies, as he suggests that Muslim Gypsies perceive the use of Romanes as essentially

Christian and avoid using it as a result (1860-3: 42). Paspates' work points to an older dialectical

diversity that is echoed in modem Turkey's Gypsy communities, and an interesting explanation as

to the difference in competence noted between groups that might suggest why Romanlar who

originate in the mübadile may have a higher competence than those located in Turkey prior to this.

In an interview with an old Romanlar woman in the Bari (Peace) mahallesi in Mersin (Icel) in May

2007, she described the 1930's migration she had experienced from Selanik (Saloniki) and the

process of conversion from Greek Orthodox Christianity to Islam at a point around Diyarbakir, with

the shift from being Greek-speakers to becoming Turkish-speaking as they become incorporated

into the new Turkish nation-state.

Other writers who referred to the Gypsies of Turkey included Francis Hindes Grooome for the Chambers Journal in 1878, whilst Dr. A. Elysseeff briefly described the "Gypsies of Asia Minor" in

his 1889, Journal of the Gypsy Lore Society article. More extensive descriptions come from Sir

William Reginald Halliday (1922: 163 - 189) who depicted the inter-dependence of certain nomadic Gypsy groups and Yörüks in the Anatolian plateau, whilst G. L. Lewis (1955) building upon the

work of Halliday, suggested that some of the latter, the Geygelli were indeed Gypsies rather than Turkmen as they spoke a variety of Romanes. Burr, in 1948 attempted to investigate the historical

situation of one particular group of Gypsies during Ottoman times (executioners; Burr, 1948: 78-79),

a position that would appear to have lasted into the modem Republican period, as Nazim Hikmet

wrote, "... if some poor Gypsy's hairy black/spidery hand/slips a noose/around my neck... " in his Letter To My Wife (November 1933). Juliette de Bairacli Levy (1952: 5-13) also provided a description of Istanbul's Gypsies, based upon her earlier travels (elaborated further in her 1953 book), in particular those musicians and dancers of Sulukule. Others had noted the Dom Gypsies of the Ottoman Empire's Arab lands, such as Father Anastäs, the Carmelite (1913-14: 298-320), who gave a detailed description of the lives of these itinerant metalworkers and traders. Robert

46


Alexander Stewart Macalister also wrote of the Nawar or Zutt in this period, though concentrating

upon their language (Domar), which he noted as maintaining a third neuter gender and therefore

being related to, but separated from Romani and indicative of an earlier migration from India by the

Dom (Macalister, 1909: 120 - 126,298 - 317; 1912: 289 - 305; 1913: 161 - 240). Both the Carmelite

priest and Macalister noted the considerable degree of prejudice shown towards the Dom by the

Arab population just on the eve of the explosion of Arab nationalism in the revolt against Ottoman

rule of 1915.

Much of the scholarship from the late nineteenth and early twentieth centuries continued to focus

upon the paths lain out by Paspati, namely language and origins. Other Gypsylorists (a term coined

by Acton to encapsulate the work of the members of the Gypsy Lore Society and other

organisations or individuals associated with this predominantly folkloric approach; Acton, 1974),

such as ̀ Petulengro' (Bernard Gilliat-Smth) chose to focus upon the customs and culture of Gypsies

in the European provinces of the quondam Empire (1915-16: 1-28,65-109) attempting to capture

something of the occupational and cultural distinctions between groups in the rapidly changing

post-Ottoman Gypsy populations of the Balkans, many of whom would migrate or be forced to do

so in the inter-war years. For these researchers, the continued pattern of `traditional' trades and

occupations amongst these groups (see also Gjorgjevic, 1929: 7-25; Byhan, 1908-09: 45-50; F.

Foster, 1935-36, for examples) of Ottoman Gypsies had preserved cultural and linguistic forms that

were not present in the rest of Europe. The dominant ethos of Euro-centric anthropological and folkloric research that Acton critiques, with its basis in scientific racism and post-Darwinian

taxonomies is to be found in many of these works, reflecting wider Orientalist notions of the "Turk"

in general that had come to permeate the discourse around the "sick man of Europe" for the

previous century, and clearly indicated by the work of scholars such as E. H. Palmer (1867, see below). Yaron Matras has criticised Acton's use of the term as a "shorthand" for dismissing this

body of work as racist in conception and therefore of little worth, and of the term itself becoming a

pejorative applied to opponents of the modem kshatriyya theory (during the debates regarding Romani origins at "The Other Side of Europe", the Second International Romani Studies

Conference in Istanbul in May 2005, at Istanbul Bilgi University), without much regard for the

accuracy or veracity of such a critique. In the context of this book, the term is used by me as a description for a particular body of texts that are associated with the Gypsy Lore Society past and

present and in some senses provide what might be seen as a romantically emplotted, synecdoche trope in the historiography of Romani Studies (White, 1973), or an identifiable narrative within this

47


historiography.

The history of the Gypsies in the Ottoman lands and successor states has been a subject for enquiry

from surprisingly early on. In attempting to explain the organisation of the Ottoman State and its

phenomenal ability to prosecute war, European commentators often examined the Empire's peoples

as part of the assessment. Aaron Hill (1685 - 1750), an inveterate traveller and observer in the

eastern lands, in his first history of the Empire (1709) takes up the notion of the Egyptian origin of

the Gypsies, albeit with an unusual twist (chap. 25). When relating the conquest of Egypt by Selim

Yavuz (the `Grim') in 1517, Hill suggests that the defeat of the Mamlüks left a rebel band led by

one Zinganus, who continued to resist the Ottoman Sultän and his son, Suleiman Kannnuni (the

Lawgiver, whom we know in western sources as ̀ the Magnificent' 1521 CE - 1566 CE). This band

captured women in the desert and expanded through others seeking to join, thereby a society came

into being of these Zingania and in them we find the origin of the Gypsies (in his 1740 edition,

chap. 25). John Hoyland of Sheffield (1816) also drawing upon Grellman (1787) mentions the

origins of the Gypsies amongst the rebels against Selim's rule, led by one Zinganeus (1816: 116).

Ottoman commentators on Gypsies in the Empire included the great Evliya celebi (c. 1611-1679),

referred to above, who recorded events and descriptions of the empire in the 17th century (partially

translated in 1834 - 50), who notes their presence upon a number of occasions in the course of his

travels in Izmir, Istanbul, Bitlis, Diayarbakir and in Rumelia, or European Turkey (though he does

not draw any distinction between those in the eastern part of the Empire - presumably Dom or Lom

- and those in the western provinces). Apart from tax records or defters (Marushiakova and Popov,

2001: 26 - 28), there are occasional glimpses in some Sultan's edicts or firmans (Fraser 19921: 72 - 175; Hasluck, 1948: 1-12). As one would expect, the concern of Ottoman elites for nomadic and

semi-nomadic populations including Gypsies, is limited to those points where the state

accommodated, conscripted, taxed, attempted to control and settle, or came into conflict with these

groups (Marushiakova and Popov, 2001: 57-58). European discussions of the Gypsies in the

Ottoman Empire (this formula is almost always used) begin to become widespread at about the

same time as European domination of Ottoman markets and the Ottoman Empire undergoes a series

of reforms arguably inspired or influenced by Western models (see Heyd, 1961).

The promulgation of Romantic nationalism amongst the various Balkan peoples, from the first

rebellion of the Serbs in 1804, the Greek war for independence, 1821 - 26, and the establishment of the Bulgarian state in 1878, to the final collapse of the Empire in the aftermath of the 1914-18 war (see McCarthy, 2001; Levy, 1975), is almost coterminous with the development of Gypsy Lorism,

48


from Hoyland (1816) through James Crabb (1831), George Henry Borrow's various works (1841;

1851; 1857; 1888) and those of Francis Hindes Groome (1881, who is frequently overlooked in his

contribution to the early Gypsy lore scholars; see for example his paper to the International Folklore

Conference of 1891, where he developed his original Encyclopaedia Britannica article of 1879 on

the influence of the Gypsies in transmitting folk tales from their Indian origins to Europe; see Owen

Jones, 1967: 71 - 80). The foundation of the Gypsy Lore Society in July 1888 (of which Groome's

was a active member though initially reticent to take any official post) and its concomitant

publication, the Journal of the Gypsy Lore Society (1888-) was an attempt. at co-ordination of

research and knowledge about these people, much of it concentrating upon the Gypsy communities

in south east and eastern Europe. Earlier published examples of travellers' descriptions of Gypsy

communities in the region (Bright, 1818), set much of the timbre of the later contributions to the

Journal of the Gypsy Lore Society. Until the advent of the work of Margaret Hasluck, the studies of

ethnographers, anthropologists and folklorists was built upon the prevailing notions that dominated

views about the Gypsies in Europe, with little recourse to those in the Ottoman lands (Paspati's

early contribution to the Journal of the Gypsy Lore Society notwithstanding). Hasluck, it has been

argued "... can be regarded as an innovator who insisted on an Ottoman discourse of the Gypsies.

She used for the first time, through her Albanian contacts the available Ottoman documents, to

reconstruct the Gypsies' experiences during the Ottoman period. " (Ginio, 2000). Her use of some of

these documents is however, not uncontested by some Ottomanists who consider her translations of

these firmans to be inexpert (Ginio, 2000).

In the post-Ottoman period, very little has been written about Gypsies in any of the Islamic

successor states of the Middle East (see an example in Hanna, 1982; 1993) with what there is

sometimes being disputed and incomplete (Lockwood, 1978: 304-308, where the author includes no information about Shi'ia Gypsy groups in Iran or Alevi Gypsies in Anatolia). In the south eastern European context, Gypsies appear most often in nationalist literature as stereotypical images or

carriers of myths (Andric 1994, for example where the Gypsies are presented as the `willing'

executioners of the barbaric Turks), or as tropes for particular qualities of other ethnic groups (see

van der Port 1998: 8-10,133-176). After the work of Halliday, there is more interest in the historical

past of Turkish Gypsies (Hasluck, 1938: 49-61; 1948: 1-12) though the linguistic and cultural aspects

of Turkish Gypsies also continue to be discussed by some authors, though somewhat incompletely

(Arnold, 1967: 105-122). It would seem to be the case that Arnold's observations of Turkish Gypsies

were conceived upon a series of notions of mobility and nomadism, despite clear evidence of Gypsy

49


settlements since the Ottoman period (Arnold suggests that there are no Gypsies to be found

between the Istanbul suburb of Üsküdar and Ankara, and does not record any of the large numbers

of nomadic Dom or Alevi Rom to be found in the region by others like Halliday, as he travels

between Kayseri and the Iranian border; 67: 106). Arnold places an emphasis upon linguistic

identification in his study, defining Gypsies as speakers of Romanes, a distinction that ultimately leads him to conclude that the European Roma are both proficient and `pure', whilst the Gypsies of

Turkey and Iran merely use what he describes as "secret languages" or cant. This situation is the

result of loss of contact between the Asiatic and European Gypsies, in Arnold's view (1967: 119-20).

Paspati's suggestion that the majority of Turkish Gypsies were `nomadic' seems to have influenced

Arnold profoundly, without considering the evidence to the contrary, and the perspective that he

adopts (the loss of Romanes for Gypsy communities between the Indus to the Bosporus) is strongly indicative of the continuing influence of 19`h century folklorism in Romani Studies. This influence

and Arnold's ideas in particular, may also explain to some degree the lack of concern by European

scholars for the study of Gypsies in Turkey in the mid-twentieth century, until relatively recently.

In many ways the scholarship of this period also continued to reflect the 19`h century concern with Orientalised models of Gypsy identity, and racially determined concerns of inheritance, `blood

lines' and physiognomy (Lee, 2000), in common with those applied to late Ottoman and early Turkish society generally (cirakman, 2002: 25-30). Yet, with the focus upon the European Roma in

the years directly following the cataclysmic events of 1933 - 1945 and the Nazi regime (see

Burleigh and Wipperman, 1991; Burleigh, 2000), the widespread knowledge of the conditions

pertaining to Turkish Gypsies seems to have ultimately slipped from the purview of scholarly

research. Arnold, lengthy article, and Mozes F. Heinschink's studies of the dialect of the sepetciler in Izmir (Cech and Heinschink, 1999) represent a small number of works regarding Gypsies in

Turkey, and in the east in general produced in this period by scholars. The devastation of the 1939 - 1945 conflict seemingly resulted in an intellectual shift back to Europe, perhaps to attempt an answer to the horrors of fascism and racial biology; whatever the cause, the concentration upon European and western European Romani communities is marked in the immediate post-war era and well into the 1960's in the Journal of the Gypsy Lore Society, and Romani Studies (a refashioning of the Journal of the Gypsy Lore Society in the face of a more critical, analytical and reflective sociological and historical approaches) scholarship in general. Turkish research remained relatively unconnected to this wider Romani Studies discipline, as it evolved in the post-war period, until the advent of the Istanbul Romani Studies Network (2002) and a series of seminars with two

50


international conferences during 2002-2005, that began to bring scholars from Bulgaria, Sweden,

the United States and the United Kingdom into contact with those who had been working largely

isolated from each other in universities, as journalists and a few individual research students.

In terms of the Turkish material produced prior to this point, that of Dr. Ali Rafet Ozcan has been

particularly influential as it was the output of a researcher from an important university, itself

supporting that research. Dr. Özcan's monograph, Turk cingeneler was published initially in 1998

(upon which Dr. Ozcan was made Associate Professor of Theology at Gazi University, Erzerum)

and subsequently made widely available through the Ministry of Culture in 2000. The work

contained references to Gypsies as thieves, prostitutes, beggars, usurers and profiteers, profligate in

their ways and immoral in their behaviour in marriage (a subject that he goes on to develop in his

later article; 2006: 461-470). He also remarks upon what he suggests is their inherent querulous and

quarrelsome nature. Such work has also informed Turkish dictionaries and encyclopaedias produced

officially, in that many of the definitions of the term "Gypsy" contain all of these prejudices and

more (including references to Gypsies stealing children and selling them, being only half a nation - following the old adage that there are 66'/2 ̀ nations' in Turkey - and, most damaging of all in the

current climate, being without religious belief). Such works have been challenged legally on a

number of occasions by the activist and writer Mustafa Aksu but sadly with varying success (Dr. Ozcan's work has been withdrawn as a result of a legal challenge from Mustafa Aksu in 2004,

though not officially refuted and remains in the catalogue of the National Library in Ankara). Other

works have been produced by journalists such as Nazim Alpman and Sinan Sanlier, who have

attempted to present a balanced perspective and corrective to the widespread prejudices and

misinformation present in Turkish media and academia elsewhere (Alpman, 1994; and various

articles by Sinan $anlier, 2006). The current campaign to save one of Istanbul's oldest Romani

neighbourhoods at Sulukule (established in Byzantine times) has also drawn a great deal of interest

from journalists and writers elsewhere in Europe, such as Kai Strittmater from Süddeutsche

Zeitung, Ayla Albayrak at Helsinki Sanomat, Fred Taikon at t Romani Glinda, and many others. This reportage has all endeavoured to include details of the Romani community in Turkey and especially Istanbul, illuminating the social and economic conditions, and the cultural life of Turkish Gypsies, in the mainstream media. Academic research has increased in more recent times with a number of studies being undertaken by Dr. Emine Onaran Incirlioglu at Bilkent University (2006: 191-201) and Dr. Suat Kolukirik at Ege Unversity (2006: 133-40). Dr. Sonia Tamar Seeman's

research into ethnicity, identity and music has also been an important contribution to the scholarship

51


about Gypsies in Turkey, moving as it does between the notions of identity so frequently associated

with them and analysing the discourse surrounding these notions and stereotypes (2006).

Finally, the encyclopaedic material to be found regarding Gypsies of the quondam Ottoman lands is

primarily located in the Encyclopaedia of Islam, 2nd Edition, under the entry "Qingäne" (Lewis and

Quelquejay, 1983: 40b - 41a), which offers the detail about the possibility of Turkish Gypsies in

Western Anatolia sheltering under the label of "Yürük" (Lewis, 1955) and confirms the confessional

diversity amongst Turkish Gypsies by citing the same Geygelli as Alevi (though not inter-marrying

with other Alevis, as suggested by my own research in the Kustepe neighbourhood in January

2007). Mehmet Fuad Köprülüzäde's 1935 article in the Türk halk ansiklopedesi, "Abdals" also

notes this confessional diversity by first drawing attention to the Abdallar as a Gypsy group, and

Melih Duygulu provides an interesting summary of the history of Istanbul's Gypsies including a list

of neighbourhoods where (at the time) they were to be found including Kasimpaýa, Kuýtepe, Balat,

Ayvansaray, Gaziosmanpaýa, Üsküdar and of course, Sulukule. Nomadic Gypsies were to be found

in Kagithane, Merter, Ümraniye, Kücükbakkalköy, Beyköz and Halkili (where a large Alevi

community resides at present). Duygulu's article also discusses the linguistic situation, suggesting

that the use of Romani, called "Romanca" [pronounced Roman ja] was widespread amongst

Istanbul's Gypsies, though couched in Turkish grammar (1994: 514 - 516). The article by M. Tayyib

Gökbilgin, in the MEB's Islam Ansiklopededsi, is more widely known though far less neutral than

that of Duygulu (1988: 420b - 426b), drawing heavily upon older sources and H. M. G. Grellman's

Die Zigeuner (Leipzig, 1783) in particular. The MEB is the National Board for Education,

indicating why this article should be more widely-known as it is in every school and education

establishment as a basic text. Far from exhaustive, this brief survey illustrates the threads of

scholarship that may be said to relate to Turkish Gypsies.

The study of minorities in the Ottoman Empire and Turkey has moved away from the predominant

concern with the Christian and Jewish minorities that were the primary focus of European, Russian

and American interests in the late 19th and early 20th centuries (see Braude and Lewis, 1982) and have remained so for scholars looking to establish the `legitimacy' of a tolerant, pluralist

commonwealth (Lewis, 2002 "Aspects of Change": 323-489) , to more concern with other groups

including Muslim minorities (see for example Neuberger, 2004; Destani 2007). The focus upon

religion might still said to be at the heart of many scholars' enquiries, and in particular with the less

orthodox expressions of Islam and Sufism (itself a field of enquiry that has been the focus of European interest for more than a century; Nicholson, 1914; Palmer, 1867; MacDonald, 1909;

52


Brown, 1868). Interestingly, there is a confluence of interests for some scholars between the

Gypsies and the mystics of Islam; for example Edward Henry Palmer with Charles Godfrey Leland

and Janet Tuckey, published a collection of English Gypsy songs translated from the Romanes

(1875). Palmer also produced a catalogue of Turkish manuscripts in Trinity library and an important

translation of the Qur'an (see Beasant, 1883). Clearly the study of Gypsies related to the wider

notion of Orientalism that prevailed in 19`h century Europe, and in particular the study of the

Ottoman Empire (see cirakman, 2002, for a discussion on the European Orientalist approach to the

study of the Empire; see also, Grosrichard, 1998).

The impact of modem historiography has been felt in Ottoman Studies as in all other areas of

historical enquiry, and there has been a shift from the tendency to analyse the Ottoman Empire in

terms of its proximity to the model developed by Perry Anderson (1979) of the Asiatic Mode of

Production to a richer discourse (Faroqhi, 1999: 212), to one that reinterprets the Empire in the face

of the `traditional view' of dissolution and decline in the presence of European ideologies and

economic power (see Aksan and Goffman 2007: 4-18). The notion that the Empire was a unique

phenomenon, a sui generis political entity monolithic and incomparable to the (equally monolithic)

polity and culture of Europe has foundered upon the studies that have recognised both external

convergence and profound internal differences (Aksan and Goffman, 2007: 21; Barkey, 1994: 230-

35). Challenges to the accepted and unsatisfactory dictums about Ottoman social structures and

societal organisation of the earlier Empire began to be challenged in the early 1980's (Braude,

1983: 69), a development imperfectly followed by some Romani Studies scholarship (Barany

2002: 53-57; Barany 2001: 50 - 63), though the notion that "the Ottomans devised a sociopolitical

system that was humane and tolerant" (Barany, 2002: 84) remains strongly embedded, despite the

significant challenges to the notion of a "humane and tolerant" pluralist polity, from Bat Ye'or and

others (2001). References to the Ottoman Empire retain the notion of a tolerance, or at least a benign indifference to the Gypsy populations, such as that by Fraser (1992: 175), a picture that is

more carefully considered in the work of Marushiakova and Popov (2001) and revealed by the

study of Ottoman legal registers in Salonika, by Ginio (2004), clearly demonstrating that Gypsies

occupied a secondary and frequently marginal place in Ottoman society that was based upon both

notions of ethnic and religious identity, something that Makdisi makes clear in relation to Ottoman

social organisation and too frequently overlooked, when he notes the "classical Ottoman imperial

paradigm [was]... based on a hierarchical system of subordination along religious, class and ethnic lines" (Makdisi, 2002: 768).

53


The study of Ottoman mentalites has produced some recent scholarship that has focused closely

upon the ideas behind Ottoman superiority and ultimately self-destructive notions of identity, with

tragic consequences for the Empire's minorities (Reid, 2000: 175-189; Deringil, 1998). Definitions

of osmanlilik (the quality of being Ottoman) altered fundamentally over the course of the nineteenth

century, culminating in the notion of the Empire as an Ottoman Turkish polity, with the modernised,

reforming Istanbul elite spatially and temporally distanced from the host of purportedly unreformed,

"stagnant ethnic and national groups", the minorities that made up the subjects of the Sultan

(Makdisi, 2004: 769 - 771). The examination of this process of ideological legitimation,

modernisation and the development of "Ottoman Orientalism" has provided a revision of the reign

of Abdul Hamid II (1876-1908) in particular, and an indication of the mechanisms underlying the

shift in perceptions of those minorities that previously were seen as integral, if differentiated

elements of the Sublime State, the religious and ethnic minorities of the Ottoman commonwealth

(Göcek, 1996; Derengil, 1998; Karpat, 2001). Models positing a culturally homogenous elite

(Miller, 1941; Lybyer, 1913), have provided us with details about the Empire's rulers and ruled in

particular periods, but too often been taken to apply throughout the Empire's lifetime unchallenged,

undifferentiated and unchanged. More sophisticated analyses have allowed the picture of the

Ottoman umerä (the religious scholars, judges and imams), and asked classes to emerge as a

dynamic force in the Empire, whose origins and extended networks of patronage and clientage

crossed religious, class and frequently ethnic boundaries (Kunt, 1983 for example). The

transformation of these classes and to some extent the relationships that characterised it, have been

identified as a key element of the altered relations between governors and governed in the later 17th

century (Kunt, 1983: 95-99), and the effectiveness of the Ottoman bureaucracy from the 18th to the

20t" centuries (Findley, 1980). As an aspect of reform, the charting of these changes in the Ottoman

Empire can be related to the study of Gypsies in the context of their impact upon them. The

significant changes that took place in the Empire, especially after Sultan Mahmüd II (1808 - 1839)

in the period that has come to be known as the Tanzimat, or `re-ordering' of the Empire (Shaw,

1977), were such that a profound shift in metalites took place that reached all aspects of Ottoman

society. As Marushiakova and Popov note, "the process of reform (or at least attempts at such) in

the Ottoman Empire also affected the Gypsies", in the government's desire to regulate the civil

status of the Gypsies in line with other groups in Ottoman society (2001: 57 - 59), something we can

see in the early reference from Paspati to the changes in the law regarding military service and Muslim Gypsies (Paspates, 1888: 3-5).

54


The fragmentation of the elites and power in 17th and 18th century Ottoman society, led to the

emergence of those known as aQyän (notables) or dere-beys (valley lords) and other rebellious

groups of irregulars such as derbenci (Yörüks), martolos (Greeks), levends (Kurds) and efläks

(Bosian Muslims), the last of which used a dialect that would appear to be drawn from Romanes,

according to Evliya celebi (Karpat, 1972: 243-281; Shaw, 1971; Fleming, 1999; for the language of

rebellious efläks see Lazarrescu-Zoban, 1983: 307-330), and to the development of strategies for

state centralisation by the sultans (Barkey, 1994). Analysing this group has led to a complex picture

of seventeenth and eighteenth century contests for resources and control, washing back and forth

between centre and periphery sometimes to an almost mortal degree, as during the reign of pro-

reform Sultan Selim III (McGowan, 1981; Levy 1975). The role of the aQyän in the history of the

Ottoman Gypsies in general is important, and I shall return to it in a later chapter, but the impact of

individual aQyän in reform, and their eventual incorporation into the process of reform is one of the

themes of the tanzimat period in Ottoman history (Barkey, 1994; McGowan, 1994: 637 -757), which

follows the promulgation of the Gülhane Rescript in 1839, and is usually deemed to finish with the

accession of Abdul Hamid 11(1876). The debate on the origins of the tanzimat in Western European

drives for reform and (ultimately) control over the Empire is also one that has been contested (Davidson, 1990; Abou-El-Haj, 2005), with those seeking indigenous models for the initiatives

constructing a well-argued, if contested case (see Darling's 1993 review of Abou-El-Haj: 118-20).

The influence of the purportedly arch-conservative religious theocracy the ulemä as a force for

change, has been cited (Gawrych, 1987: 91 - 114; Heyd, 1961: 63-96), albeit in the form of the

leaders of the less orthodox Mevlevi dervish order, founded by Djäläl al-Din Rümi (known as Mawlana, or `our master', b. 1207 - d. 1273).

Charting the `decline', `demise' or `collapse' of the Ottoman Empire is an exercise almost as old as the Empire itself, not an exclusively European affair (Lewis, 1962: 71-87) and one that has been

significantly challenged by recent scholarship in Ottoman Studies (see Tezcan, 2007: 167 - 198).

The usual reasons for such decline, many of which reflect the cyclical book of Ibn Khaldun's

Muqqadimah, frequently fail to establish the actual conditions in the empire at the time or identify

the major causal factors, relying instead upon European notions most pithily expressed by Sir Paul Rycaut, who observed that though not entirely fallen, "... the Empire may never rise again" (1682: 5). Bruce McGowan's painstaking and perceptive study of Ottoman economic life in the Central Lands (1981), clearly demonstrates the underlying demographic collapse in the wake of extended warfare, and the consequent near-bankruptcy of the Ottoman exchequer with consequent

55


loss of revenues, supplying an alternative narrative to the consistent discourse of Ottoman viziers

and courtiers who portrayed the ills of the Empire as stemming from a variety of less structural

causes, not least the interference of women in its governance (Pierce, 1993). Michael Palairet's

analysis of the Balkan economies in the final period of the empire, demonstrates the continuing

impact of Ottoman economic structures upon the region in the period of greatest European market

penetration (1997). Donald Quataert (1983) has written extensively, documenting the profound

economic dislocation resulting from internal revolt and rebellion had upon Great Power - Ottoman

relations, whilst Kenneth Cuno has challenged the nationalist rhetoric of a sharp separation between

the Ottoman and `modern' periods and the impact of European economic penetration of agrarian

relations in Egypt under Muhammad Ali (1805-1848), for a longer term assessment of the

developments of cash-crop farming, commodification of the land and the emiseration of the

fellaheen (1992), of which the Ghagar, Nawar and Halebi (all terms for Gypsies in Egypt, that are

primarily used by non-Gypsies, who prefer the term Dom; see Marsh, 2000) formed a dependent

part (Hanna, 1982). Economic dislocation also formed part of the process of disruptive change for

the Gypsies of Turkey and the Balkan lands, as I shall go on to argue in this book.

The key to the 19`h century Ottoman state was its success in maintaining an accommodation

between the needs of the reformed bureaucracy, European political - economic interests and

diminishing territorial possessions. This complex balancing act was only possible when access to

markets could be guaranteed, the Ottoman economy effectively dominated by Europe and the

Ottoman governance of the Empire maintained in the interests of the `Great Game' (see Ufford,

2007; Hopkirk, 1990), or the `Game of Nations' as it was known (Copeland, 1969). It was never in

the interests of Europe's Great Powers to undermine Ottoman control of the Empire, except when

disruption to markets became acute and access could be ensured with emergent successor states

(Quataert and Inalcik, 1994: 761). Quataert's analysis illuminates the relationship between European

sponsorship of nationalist movements in the Balkans and exploitation of raw materials in return for

finished goods. Nationalist historiography in the region prefers to gloss over the more material

interests of European government's support for freedom and liberty for some of the Empire's

minorities (Dimitras, 2000: 41-59; Sesic, 2005). The examination of the situation of non-Christian

minorities in the Arab lands has received less attention from historians and ethnologists (Destani

being a notable exception), as has much comparative work with other imperial systems (Barany's

analysis of the Ottoman and Habsburg Empires, coming as it does from a combination of political

science and Romani Studies is unusual; 2002: 83-110). This book draws upon a number of differing

56


literatures in order to suggest a series of conclusions regarding Gypsies, and the focus here has been

to survey the salient debates within the field of Ottoman Studies and Ottoman minorities in

particular. The shift in the position of Ottoman Gypsies is related to the broader changes within the

Empire, often resulting from European influences that altered the relationship between state and

minorities drastically in the later period.

One other consideration that emerged from European-Ottoman relations was the influence of time,

as measured in European terms. Timepieces had been presented to the Ottoman sultäns by European

diplomatic missions as early as the 16th century, but the general reform of time regulation was a

product of the Tanzimat period of the mid- 19th century, when the Empire began to be governed by

clock-towers and the hours of the day were no longer determined principally by the call of the

muezzin. Concomitant with notions of progress, the acquisition of timepieces for the elites in

Ottoman society became a symbol of modernism, of adherence to the programmes of reform and a

marker between the boundaries of civilised and uncivilised. Pocket watches had a particular

resonance, not only in terms of status, but also in semiotic terms - as Ahmet Hamdi Tanpinar has

written,

"The watch which is the most intimate friend of its owner, companion to the beat of his

pulse at his wrist, a friend at his breast sharing all his joys and sorrows, heated by the

warmth of his body and espoused by his organism, and the clock which stands on his table

and experiences with him all the happenings of the time span which we call a day, both

inevitably come to resemble their owner, and become accustomed to think and live like

him... " (1961: 46)

Adoption of the western European notions of time had profound impacts upon Ottoman sensibilities

and conceptions of `progress' and `evolution', ones that maintained the differences between those

modern populations and those that could not be modernised, or Europeanised (Makdisi, 2002: 771- 3), with Istanbul as the temporal centre of a coeval and equivalent Muslim civilisation that resisted European imperialist and colonialist dominance in temporal and spatial terms (Makdisi, 2002: 771; Fabian, 1983: 144). It is in the context of these notions of time that the history of the Gypsies has

come to be written.

57


1.4 Gypsies, history and the problem of time

"Time is the element in which we exist. ... We are either borne along by it or drowned in it. " Joyce Carol Oates, Marya:

A Life (1998)

TIME, AS JOHANNES FABIAN HAS CONVINCINGLY ARGUED is an ideology, a means of exercising

power over those who were and are different, made distant in space and time (Fabian, 1983: 144-

47), and thus colonialism is justified through the bringing of time to those without it in the name of

progress, advancement and development (Makdisi, 2002: 771). European time is a continuous flow,

a trajectory along which the less advanced societies and peoples are marked and mapped, always in

relation to the `standard time' of the western world, denied an equivalence and the right to be

coterminous, but different (Fabian, 1983: 145). The issue of time, and its construction in the writing

of history, anthropology and ethnology has been one that has been little focussed upon in the field

or discipline of Romani Studies. Outside of the title of Michael Stewart's monograph and the

English title of Emir Kusturica's popular film (1997; 1988), the notion of time and the Gypsies has

rarely been discussed (see the recent initiative by the Open Society Institute Roma Initiatives,

2007). There are of course notions of time implicit or explicit in the collections of Gypsy folk-tales

that many of the early Gypsy Lore Society members gathered and published in their Journal... and those of Francis Hindes Groome (1899), but these are the references to the timelessness that all such tales exist in, using the tropes of "From the dimmest dawn of time... " for example, the opening of "The Dead Man's Gratitutde", Turkish Gypsy Stories, no. 1 (Hindes Groome, 1899: 1; an interesting

version of the Tobias and the Archangel Raphael story, from the Book of Tobit, and based in

Zoroastrian myth - see Vickers, 2000: 339-342).

All the historical writing that we might define as comprising Romani historiography has been done

so in the framework of western European notions of time, sequence and events, the basis of which it has been argued is to be found in the Neolithic impulse, the dramatic metaphysical shift from

cyclical measurement to linear progression (Martin, 1993: 53 - 74). Defined, classified, measured and documented in notions of sacred (Judeo-Christian, Muslim) or secular (post-Darwinian) time, the study of the Gypsies has been bounded by the underlying conceptions of how they correspond or differ as a group, to non-Gypsy structures including the quantification and management of the sequence of events that has come to predominate over all others. In effect the construction, or reconstruction of the Gypsy past has resulted in a great deal the reflects the present (whichever

present the author occupied or occupies; see Burton, 1988: 420 - 433). The "time of the Gypsies" is in fact the allochronic time (placing the object of study in another, non-European time, as opposed

58


to acknowledging the contemporaneous relationships of power and inequality that exist between

researcher and researched), of those who observe and then write about them and in this the distance

and separateness that Fabian describes as a product of European scholarship, is marked out in non-

Gypsy terms (Fabian, 1983: 144-47). The question here is one of how the imposition of non-Gypsy

time has impacted upon the Gypsies, and following that, whether it is important and necessary to

both recognise this fundamental organising principle and ask whether there is a particular Gypsy

metaphysic that is being distorted through this process of imposition of western European time and

scholarship.

The origins of notions of time, as suggested above, have been argued to lie in the shifting

metaphysic of hunter-gatherer societies of the Palaeolithic to the agrarian sedentarism of the

Neolithic era, when the management of cycles of oestrus and birth in livestock, and sowing and

reaping in agriculture became the focus of communities hitherto adapted to these, rather than

adapting them to their needs. The measurement of these cycles saw the consequent development of

surpluses, population growth, elite groups who extracted those surpluses through co-option (the

priesthood) or coercion (the nobility and kings or queens), stratified societies and highly complex

social organisation. The control of time became the prerogative of the priesthood and connected to

the measurement of dynastic achievements, changes and legitimacy through association with the

past - the establishment of history in fact. The growth and control of trade and war by the dynasties

themselves established the principle by which the arch of time was firmly established in a

relationship with dynasties, priesthood and gods, and forged into a trajectory that flattened the cycle

of pre-neolithic hunter-gatherer communities into the vector of sacred history (Martin, 1993: 56-60;

Fabian, 1983: 36 - 76). The creation of sacred time and history, as an aspect of monumentality, literally of `marking time' is intimately tied to the notion of a recording God's interventions in the

Creation through chosen individuals (prophets) or peoples (Jews, Muslims, Christians) in a divine-

human dialogue that both projects the trajectory of dynastic and priestly time forward into an

eschatological future and an immanent, deliberate point of origin. The notion of order is essential,

and the use of time in organising this narrative is the prime means of creating it. Out of chaos and disorder, divinely created or inspired harmony and concord emerges, usually through the

mechanism of language, the word as understood by humankind, or anthropologos (a distinctly

different narrative to that of many where the flux and fizz of creation is constantly re-presented in

the present through ritual and myth; see Leeuw, 1938: 413-14). The shift from sacred to secular in

the Enlightenment altered the narrative to one of progress (Providence and Manifest Destiny; 0'

59


Sullivan, 1839: 426-30), and saw the de-coupling of time from sacrality, its naturalisation and

expansion in the work of Darwin and Wallace (1958) and Charles Lyell (1835-5). The cosmic

trajectory of humankind as the recipient of a divinely-ordained mission, whether Judeo-Christian or

Muslim, had been initially harnessed to, and then supplanted by the superiority of the species over

all others, the `natural' processes of selection in the linear, branching advance over aeons of

`mankind' (the masculine gender linked inextricably with notions of order, harmony and progress in

the European imagination). Such a version of time followed the twisting, branching schema of the

sacral path that had been mapped by the prophets for God's chosen peoples, only in this the

imperative was biological rather than celestial. Not only human beings, but their extraordinary

ability to engineer space and matter, substance and sound in language and artifice, came to be seen

as part of this narrative of selection, of separation, destruction and survival. Culture and expression

subsumed the metaphysics of religion and ritual, history was extruded from ritual's dark twin, myth.

The schematic presentation of cultures and languages in this way, in a reflection of the biological

narrative of adaptation and evolution and a historical consciousness of conquest and expansion,

revealed the vector of mankind's march through time uninterrupted (Martin, 1979: 153-159). In this

narrative, the construction of time was inextricably bound to the march of industry, science,

technology (including the primary inspiration for development, war; see Diamond, 1997; Zinsser,

1935) and ̀ civilisation' (or the management of post-war societies).

The insertion of the Gypsies into this structure of time, as a means of recording the distance

between one event and another and between non-Gypsies and Gypsies, has profoundly influenced

the shape of Romani Studies. In this sense, the events that might be characterised as important in

the history of the Gypsy peoples are perceived and depicted within the framework of `an-other'

time, one that has itself been filtered through a number of cultural and intellectual `screens' that

serve to distance the observer from the observed. The narrative of Rom, Dom and Lom is measured

primarily against the western European societies that have been predominant in analysing,

categorising and describing them, in ways that frequently conform to the biological, evolutionary trajectory of post-Enlightenment, Darwinian time. Thus the movements of these various and diverse

peoples are captured within the narrative of journey that is familiar to Judeo-Christian cultures, the Biblical Exodus, and are rendered as one, undifferentiated, in a trope that is resonant to us as the

ethnic narrative of the renascent nation. In an irony that cannot escape notice, the groups of those

most dispossessed by the very notion of the nation, the ethnie and its homogeneity are encompassed in a discourse of time that defines these very limits of belonging, group and history. The

60


construction of such narratives is one of the primary motivations behind all history, the desire to

narrativise and chronicle the fragments of the past (White, 1966). In this desire, the element of time,

as conceived of in colonial terms is a key factor, the drive to categorise, to create taxonomies of

time and peoples.

The primary relationship with time for the ancestors of the Romani people, those captives of the

Amir (later Sultän) Mahmüd who were forcibly removed to Ghaznä in the late 10th and early 11'h

centuries, must surely have been with what we might describe as a collision with Muslim time, the

period since the Blessed Prophet had taken the momentous step in the migrations to Madina with

his Companions, or at Hajj (AH). Their experience as ghuläm (captives, soldiers and servitors) and

äbd (slaves) of the Amir and his governors elsewhere in the empire, would have brought them into

the orbit of the redemptive trajectory of Islamic eschatology, and the influence that Persian thought

and sufi mysticism had by then had on the Islam of the region (see, Richard N. Frye, 1974). The

complex nature of Islam in these lands was reflected in the person of Mahmüd himself (see below)

who would appear to have been quite different to the ideological picture drawn of him in the

chronicles and panegyrics of his reign as the pre-eminent ghazi warrior of his day, and in the

responses to the Ghaznävid invasions of India. As Romila Thapar has demonstrated, the notion that

these can be situtated in a discourse of conquest and resistance, is part of an epic narrative in the

colonial and post-colonial periods (Thapar, 2000a: 46), where the dichotomy between the Hindutva

and Islamist discourses as archetypal conflicts becomes crystallised. This picture of implacable

hostility is a product of modem times, as is the communalism that Usaama Makdisi has described in

Lebanon, or Gyan Pandey in India (Makdisi, 2000: 180-208; Pandey, 1992). As such, these

narratives are a product of the colonial time imposed by the rulers of these lands, the Ottomans and British, and such discourses have become a part-and-parcel of the historiography of Romani

Studies, as mentioned above.

The specific case of Mahmüd of Ghaznd's invasions of India were recast by British colonial

administrators, initially as a legitimation of their historical "mission" to rule the sub-continent (a

counter-part to the ideology of "manifest destiny"), as Sir Henry Miers Elliott wrote in his preface to the collection entitled The History of India, as Told by Its Own Historians, The Muhammadan

Period (1867-77: 3)

"... and drawing auguries from the past, he [the reader]... will derive hope for the future,

that, inspired by the success which has hitherto attended our endeavours, we shall follow them up [the Muslims]... by continuous efforts to fulfil our high destiny as the rulers of

61


India... "

then in the course of confrontation between the government and opposition in the House of

Commons in 1843, as the "Hindu trauma... which has been rankling amongst the people for nearly

a thousand years.. ." (Ellenborough, 1843). The notion of the "Hindu trauma" that Ellenborough

coined has become a part of the discourse surrounding Romani origins (inherited from the Hindu

nationalism of `Veera' Savarkar and K. M. Munshi; Savarkar, 1928 and K. M. Munshi, 1951) as

suggested above. The endemic ethnic hatreds, so much a feature of British colonial history writing

concerning India, has become embedded in the Hindutva history-writing of modem textbooks

throughout the sub-continent (Deb, 2003).

The resistance to Mahmüd's incursions was mounted by both Hindü and Muslim people in India

(Thapar, 2000a: 30), as the sources clearly indicate in their exaggerated descriptions of how many

`heretics' had been slaughtered (the same number as käfirs - or infidels, fifty thousand in the

assaults upon Mansur and Multan for example), in this case the Shi'ia and Isma lis of Sind whose

places of worship Mahmüd allegedly razed. In the chronology of the epic of conquest that 14th

century Islamic writers such as Ferishta and Isämi penned (primarily as a means of promulgating

the model of ideal kingship to Muslim rulers at the time, not least the Saldjüks; see Hardy,

1960: 107-8), Mahmüd was `fashioned' as the protector of Islam through both the institution of the

shari'a and the destruction of heretics (Shi'ia and Ismaili), thus combining the twin pillars of

Muslim rulership (Hardy, 1960: 110). Modern Romani Studies scholarship continues to accede to

these notions through an uncritical and one-sided reading of the sources, where Mahmüd's

campaigns are consistently represented as a crusade against Hindü India in order to establish Islam

upon the sub-continent, despite the obvious point that Islam had first been brought to India with

mercantile communities in the early years of Muslim expansion in the 7th century CE/ I" century

AH, and subsequently more forcibly established in the three years of Arab invasion of the early 8th

century under Muhammad b. al-Käsim al-Thakafi (711 CE/93 AH). The attempt to consolidate

Arab Muslim control under Khalif Umar II (717 - 720 CE) failed, although Muslim rulers

dominated the region and eventually allied with the Fatamid Ismaili dynasty in the 10`h century 985

CE/ 3`d century AH. The epic narrative of conquest, was in fact a re-imposition of Sunni control by

a powerful ruler allied to the Khalif'a of the time, and more importantly perhaps, almost entirely

driven by economic concerns - the means of support for an enormous standing army and militarised

state, the desire to control trade in a number of commodities, most profitably horses (an essential

element of Mahmüd's own forces and hugely profitable for the Muslim and Hindü merchants of the

62


Sind and Gujurat; see Chakravarti, 1991: 159 - 182), a trade that competed with that through north

western India and Afghanistan that was under the control of the Ghaznävids (see Chakravarti,

1999: 194 - 211). The chronology and narrative of conquest and resistance must be carefully

analysed in order not to subscribe to overblown notions of Mahmüd's invasions as crusades against

Hindus, wreaking terrible destruction and killing and capturing hundreds of thousands - even his

celebrated (or vilified) attack upon the temple of Somanätha was clearly recast by both later Muslim

and Hindu authors, though with very different intents. The actual `destruction' of the temple and the

famous ̀ lingam' is extremely doubtful (Thapar, 2000: 38), and the infamous episode of the gates, so

vigorously argued for by Lord Ellenborough proved a chimera (Brigadier Nott of the Afghan Army

and the 43`d Bengal Native Infantry [subsequently the 9t" Jat Light Infantry] returned them to the

temple on the orders of the Governor General Lord Ellenborough, but they transpired to be of Egyptian, not Indian origin and they remained in the Red Fort at Aghra, having been wrested from

the tomb of Mahmüd and the Afghan chieftains during the infamous 1842 Afghan War; Allen,

2001: 56-60). The spurious Somanätha Gates became embedded in the narrative of British colonial

time as a result, with the Orientalising of the Afghan peoples and in the mythology of the North

West Frontier. In a very real sense the discourse surrounding Romani origins is part of this

narrative, through the inheritance of Hindutva nationalist historiography, as opposed to the

panegyric of Islamist narratives of conquest and the incorporation of the British colonial epic of `civilising' India in partial forms (including the paper from Colonel John Staples Harriott, discussed

in full below, on the origins of the Romanies). Continuing efforts to slot the Roma into this

complex trajectory do so at the expense of a critical understanding of time and its socially

constructed impact upon the `Other', but perhaps more importantly serve to obscure the

complexities of the historical origins of the Gypsies and reinforce the politicised nature of these discussions, adding weight to the critique of Matras, Mayall and others (2002: 208; 2004: 243).

Their migration from the field of Dandängän in 1040 CE (see below), the battle where the Saldjüks defeated Mahmüd's son, Mas'üd, probably brought the ancestors of the Roma into contact with alternate notions of time as defined through the synchronic perceptions of sufi mysticism in Khoräsän, the most important aspect of which could be said to be the recreation of the eternal movement of the universe in the ceremony of the sema. The circular, sweeping motions of the dervi§ participants, one hand raised and one dropped (to signify `as above, so below') was conceived of as the representation of unity and wholeness -a concept still found in the oriental dance of the modern-day Romanlar in Turkey today, complete with the movements of hands and

63


feet that reflect, arguably, this pattern and the older forms that connect with Indian dance in its

wheeling motions (I am indebted to a number of Romanlax dancers in Sulukule and Kuýtepe

[Istanbul] for these insights, drawn from a series of interviews, June - September 2006). This

narrative of connection between modern dance forms and older patterns forms part of the

haeteratopic discourse of Turkish Gypsy time, an alternative trajectory or `narrative of journey' that

embeds them into a vector of Hindü-Muslim time, and one that associates them with Islam yet

offers an explanation for less orthodox practices that may be defined in the broader range of sufi

and Alevi experiences. The association of mystical Islam and the Turkish Gypsies is one that is very

much alive in the modern communities, as the final chapter will demonstrate, with groups of Alevi,

Abdallar and derviý Gypsies (historically as part of the Kalenderi order in Ottoman times; see

Lewis, 1971: 173), many of whom are peripatetic, following particular routes that take in holy

shrines or sites (like the village of Hacibekta* in central Anatolia in August, the Alevi village of

Bayraktar, Kocaeli and the environs of Halkili on the outskirts of Istanbul), and trading in small

religious items. The passage through Khoräsan on the silk routes of the southern Caspian likely

brought these proto-Rom into the orbit of mystical Islam, and the notions of timelessness, or eternal

time that pervade the various philosophies that make up sufism.

Arrival in the troubled regions of north eastern Anatolia in this period, and the lands of the

Armenian kingdoms, most especially the Baghratid Kingdom (also Bagaratid, Bagratuni in the

sources) including the Armenian lands of Syunik, Lori, Vaspurakan, Van, Taron, and Tayk (855 - 1064 CE, although the unity of the kingdom started to break up in the 10th century and the latter

period of 1045 - 1064 CE was one where the Baghratids ruled as clients of the Byzantine Emperors

in Constantinople, often in absentia as they were in the capital as virtual prisoners; Khorenatsi,

1978: 358-359). This brought the proto - Gypsy groups into contact with the miaphysite (the

doctrine of Christ's one incarnate, united divine and human nature) Armenian Apostolic Church,

one of the oldest in Christianity, established according to tradition in 301 CE. The dating used in

this church relates to the year of the 554 CE when the synod convened at Dvin formally condemned

the formula used to define orthodoxy at the Council of Chalcedon and adopted the Armenian

calendar - the Armenian Era (the Armenian Church was and is mistakenly accused of

monophysitism, the `heresy' that suggested Christ's human nature was obliterated by his divine

physis, taught by Eutyches and itself condemned by Chalcedon in 451 CE. The Eutychian doctrine

was a condemnation of the Arianism of early Nestorian beliefs, that rejected the divine nature of Christ; Hughes, 1961: chap. 4). The Armenian lands are also the likely first point of contact with

64


Christianity, aside from the possibility of chance meetings with mendicant Nestorian monks on the

silk routes, and so the primary encounter with the messianic Judeo-Christian trajectory of

eschatological history, combined in this particular case with a notion of a `chosen' people who's

destiny was to suffer, according to the timbre of many of the Armenian chronicles of the time

(Dostourian, 1993; Lastivertc'i, 1978). Armenian notions of time as expressed in the chronicles that

exist from this period demonstrate a profoundly apocalyptic version of time, in that the end of days

had been inaugurated by the arrival of the first Saldjüks inc. 1018 CE/467 AE, when they raided

the areas around Lake Van and defeated an Armenian force with their archers. The warriors were

wild-looking and demonic, according to Matthew of Edessa, with long hair that flowed like women

(A. E. Dostourian, 1993, op. cit. p. 47; the presence of women warriors amongst the Saldjük and

their Turkmen allies is well-documented in later periods, so it seems likely that Matthew's

informants shared the common abhorrence of women as combatants in identifying all the warriors

as male). Vardapet Arsitakes (Lastivertc'i, 1978: 42) suggests that the events he recorded were

clearly part of the eschatological cycle of Revelations, when he writes,

"At the beginning of... Emperor (Michael IV Paphlagon's) reign (1034-1041 CE), there was

an eclipse of the sun during the month Arac', on a Friday evening, in the year 482 of our

(Armenian) era. Many learned people, seeing... [the eclipse]... believed that the birth of the

anti-Christ had occurred on that day, or that it presaged very great evils. Indeed such...

[disasters]... did occur in our day, and this narration is leading to... [a description]... of

them. With our own eyes we saw the blows of divine anger and the unheard of punishment

directed against Armenia because of our sins. "

However, despite being especially rich in primary sources in the period when we might expect to

find some indications of groups of unusual foreigners from the east, these have not been the concern

of Romani Studies scholars by-and-large in attempts to support their linguistic arguments for the

Armenian-derived lexicon in modern Romanes. The context when such a lexicon came into

Romanes is lacking too in the secondary literature; Fraser suggests that the time spent by the

ancestors of the Gypsies - in this case the Rom and the Lom - "cannot have been brief', and further

goes on to make the point that Lomavren "shares practically no items of Armenian derivation with

European Romani' might result from an earlier separation between these two groups before

Armenian had profoundly influenced Romanes in the way it has Lomavren (Fraser, 1992: 41).

Hancock counters Fraser's suggestion by demonstrating that seventeen words are shared between

Romanes and Lomavren whilst ten are shared between Lomavren and Domart, but not Romanes

65


(Hancock, 2006: 78-9). Earlier Hancock describes thirty-four Armenian loan-words (1987: 4-10), and

indicates the context for some of them being acquired, which may give us an indication of the kind

of shift in metaphysic that I suggest had taken place. Hancock points to the changes in language that

resulted as a consequence of the contact with Christianity in the Armenian lands, a Christianity that,

as I have suggested above was almost millenarian and certainly apocalyptic at this point in time,

and may have provided a new and rich series of explanations for what was happening for the people

around the proto-Gypsy groups when they arrived in these lands as refugees and a defeated and to a

large extent, decapitated society (in the sense that the remnants of the Ghaznävid ghuläms and `abid

were no longer the organised military units and their auxiliaries, but had become melded with the

camp, the raggle-taggle assortment of those that followed the armies and others fleeing the advance

of the Turkmen).

The context for the adoption of terminology related to Christianity was perhaps one that reflected

the very real fears of ordinary people in these lands that the "end of days" had indeed arrived. The

words for the cross (trugul), church and priest (terusul) that entered Romanes and Lomavren derive

from experience of Armenian Christianity, also suggesting that the proto-Gypsy population had not

divided into those who continued to move westwards towards the Byzantine metropol and

ultimately western Europe, and those that remained (or were constrained) in the Armenian and

eastern lands (Hancock, 2002: 74). The period of time that these ancestral populations spent in these

regions was long enough and the situation critical enough, for the introduction of other words from

Armenian to become embedded in the languages of the early Gypsies, such as "Easter" and

"godfather" (Patradji, kirvo) as Hancock has pointed out (2004: 74), though this I think would have

been less of a survival strategy and much more a response to the immediate circumstances and the

need for explaining them with a new series of understandings, perhaps as a result of loss of faith in

the older notions of spiritual relationships, in a similar process to that suggested by Martin

regarding the impact of Christianity upon Ojibwa and Micmac peoples in northern America in the

17`h century (Martin, 1978). The shattering of an older metaphysic and its reconfiguration into a

new, hybrid form (56-65), influenced by the impact of subjugation to the complex Islam of the

Ghaznävids (though not, I think Islamisation, for reasons I shall outline later), encounters with

sufism in the course of traversing Khorasän and the longer exposure to Armenian Christianity,

following the defeat at Dandängän [1040 CE] and prior to the Saldjük victory at Ani [1064 CE] a

period of 20 years or so, of rapid movement and change in identity through incorporation of other

elements for these emerging Gypsies [Hancock, 2006: 70], and of significant shifts in views of the

66


world and their place in it, of loss of function [military and auxiliary] and defeat, enslavement and

destruction, both literally of the warrior-class and metaphorically of the notions of self as Hindus

and Indians). These circumstances are at the heart of the shift in metaphysic I'm suggesting took

place and the movement through space and times (Hindu, Ghaznävid, sufi timelessness and

Armenian Christian apocalypse). The notion in Fraser and others (1992: 41; Marushikova & Popov,

2001: 12), is that the context of contact with Armenian is an extended one, and not as this book

argues one that may have seen rapid change during a period of crisis and turmoil. Such times clearly

existed in these lands, and those of the Byzantines as they saw the Empire in the east collapse under

the Saldjük onslaught with their defeats in 1071 CE at Manzikert and 1176 CE at Myriocephalum.

The importance of an additional factor here is the relationship between language and war.

67


1.5 Language-change, loan-words & war

THE IMPORTANCE OF THE SHIFTS that language undergoes during wartime has been largely

unregarded in the discussions of loan words in Romanes, a surprising example of the lack of context

for the important linguistic arguments made to suggest the various narratives of journey undertaken

by the proto and early Gypsies. One of the arguments made in this book is that the speed of

language changes in the circumstances of conflict, turmoil and war are extremely rapid and the

acquisition of loan words takes place within very short time spans indeed. The wide range of

possible shifts is also one that goes well-beyond the immediate need for military terminology and

this is one of the reasons that the arguments to locate the origins of the Gypsies in a warrior class

need to be contextualised beyond the sharply linguistic definitions that are being deployed at

present, by both those who would promote and those who wish to challenge or even refute such

suggestions. This is a subject that goes beyond the purely linguistic field, into the realms of what we

might describe as socio-military history.

Linguistic forms are greatly extended during war-time, with new words and concepts coming into

existence in oral language whilst it might be said that written language contracts into abbreviations

and a vocabulary that becomes foreshortened through a codification that is intended to define

"friends" and "enemies" (Fussell, 1975: 170). In wartime, distinctions are drawn between "us" and

"them", though the contents of these can and do frequently shift and change in differing contexts,

whereby the former is imbued with what are broadly seen as morally positive qualities, notions of

being `civilised', virtuous, `just' and frequently religious, and the latter comes to constitute the

negative, barbaric and even bestial or inhuman qualities that are frequently irreligious or anti-

religious. They are cruel, uneducated, undisciplined and subject to cupidity and unnatural vices,

unrestrained in expressing their emotions, but ultimately cowardly. Frequently, they allow women a

degree of power and even allow them to participate in war as combatants, possibly as leaders

(Ascherson, 1996: 61-62). Importantly they are subjects of despotism and tyranny, non-freedom in

contradistinction to those who are fighting for freedom, democracy and against tyrants. In this way,

language becomes the boundary between who is and isn't an enemy. Language is more than a

system of communication it encompasses signs and symbols that give meaning to various

understandings, assumptions and perceptions. This whole is part of what Edith Hall has described as

"inventing the barbarian" (see Hall, 1989: 21-32). The "enemy" is often identified not by what he or

she may speak, but by what they do not speak or the fact that what they speak is unintelligible, a "babble" (barbarophonos in Greek), or cannot speak properly. Such names as Hottentot stem from

68


notions that deny a validity to others' languages, in a similar way to the term cant and reduces them

to the level of beasts of the field, those without speech (Ascherson, 1996: 64).

In the process of exclusion and inclusion, the relationship between language and place frequently

changes, especially as conflict and destruction or violence overtakes them. The names of places

alter and space becomes re-designated, such as "no man's land" (Caddick-Adams, 2007), or lose

their function as in the change from farmer's field to battlefield, countryside to "theatre of war".

The context in which people exist in their everyday lives alters too, not just for the armies and

troops but also for the civilian populations and an argot develops that attempts to encompass these

shifts and alterations, sometimes in order to distance the events and their horror - as in the

euphemisms that arise such as "gone for a Burton" originally an advertising campaign slogan for

British ale, it became synonymous with death, as did "bought it" during the First World War

(Fussell, 1975: 169-87; 1990: 251-67). Use of these argots becomes another marker of exclusion and

possibly an indication of those who are disloyal, as they do not know them. Objects become

dislocated, such as "tank", or proper nouns become generic objects, like the Roman consul Gaius

Vibius Pansa Caetronianus (German panzer). Using the "wrong" language (not knowing the argot,

idioms or responses) can be fatal - Bakongo Zairiens in Angola's bloody 1993 conflict were often

"discovered" by their assailants when they were asked to say the word arroz meaning "rice" in

Portuguese. Those who could not pronounce this word "properly" were killed as enemies (Mabeko-

Tali, 1995: 71). Those who are loyal are part of the group incorporated through language, and

conquest can result in profound language changes that are not directly related to practical necessity.

As Ottoson has argued for post-war Norway, when it comes to a situation where the potential exists

for language shifts "the most basic fact is that change is never necessary for purely system internal

reasons... " (Ottoson, 2005: 98). New meanings or variant meanings can become attached to

ordinary items and new words coined to mark the altered state in which people learn to exist.

Shared elements that originate in differing languages are common in war-time. In the Crimean War

of 1853-56, the contact between British soldiers and Ottoman troops led to loan-words becoming

part of the English language: chockablock for example from the Turkish cok kalabalik (very busy or

crowded), cakil meaning "pebble" leading to the Scottish chucky stanes, and mufti (to wear

plainclothes rather than uniform, originally Arabic). During the First World War troops from the

British Empire brought a number of these terms into common parlance, including Blighty (from the Urdu "bilyati" meaning foreign), cushy (also from Urdu, "kushi" or comfortable), shufti (from

Hindi, to reconnoitre, that also became "recce" from the French) and the experience of France and

69


vin blanc became "plonk" in the same period (Caddick-Adams, 2007).

The proposition that the proto-Gypsy groups must have spent an extended time in the Armenian

lands in order for the influence of that language to be felt in modem Romanes is, I think open to

question. The circumstances of such contact are not those that require an expended period in that

they are war-time and as such, language changes are much more rapid and transmitted quickly. The

argument that linguistic elements to be found in modem Romanes demonstrate a warrior origin (see

the discussion in section 4.1), for the ancestors of the Gypsies may also be adjusted in the light of

this argument, as the counter arguments refute the notion and the evidence as "web-page polemics"

(see Matras, 2002: 208). The suggestion here is that limiting the notion of militarily-influenced

language only to a warrior elite (though Hancock does make it clear subsequently to his earlier

work, when he writes "[that]... only a minority of the Indians would have been Rajputs in any

case... " with reference to the arrival of proto-Romani groups in Byzantium; see Hancock, 2006: 84)

does not recognise the changes and shifts possible that occur in language during war-time, with

lasting effects amongst non-combatants. I would argue here that the entire period during the

formation of the Romani people can be viewed as consisting of this "other" time - from the late 10th

century invasions of Sebüktigin and then Mahmüd of Ghaznä to the arrival of Gypsies in

Byzantium, this is an extended experience of war, conflict and turmoil that impacted upon all of the

ancestral Gypsy populations, not merely those that may have been directly engaged in prosecuting

war. The emergence of a war-time idiom amongst the population at large, of the areas invaded by

the Ghaznävids may have been further developed during the period of Ghaznävid captivity and

military enslavement, the defeat of the Ghaznävids and the rapid movement into eastern Anatolia

ahead of the Saldjüks and their Turkmen tribes peoples. The experience of the Armenian lands and

the `apocalyptic' understanding of the times recorded in the chronicles are likely to have been

another critical juncture in the development of Romanes, with rapid shifts and acquisitions of loan-

words and perhaps new, or re-allocated terms entering the proto-language, and not simply those that

reflect functional needs, as Ottosön and others have suggested above.

This is clearly the province of linguistics and as such, these arguments here may not satisfy the

expert scholarship that regards Romani Studies as primarily following Paspates' famous dictum,

quoted earlier. The historical circumstances are also unproven in that there are no clear references to

these processes as yet discovered in the sources; this is a case of drawing inferences and careful

suppositions where I may propose an alternative or illuminating narrative - but still a narrative. Here I have attempted to contextualise both some considerations about the nature of time and

70


narrative structures, and then to draw out the threads of an alternative proposition regarding the

length of time proto-Gypsy groups spent in the Armenian lands (in line with my suggestion about

the acquisition of Persian loan-words during the period of Ghaznävid captivity, in a following

section of this book) and the possible circumstances that provided the context for an Armenian

lexicon to be introduced to Romanes and shared with Lomavren. The question of who those

populations were, and are now, as they have been considered by scholarship is one that considered

here.

71


2. Who are the Gypsies? Identity and influences in Romani Studies

THE DEBATES AND DISCUSSIONS surrounding the subject of the origin and histories of the Gypsy

peoples of the world have been conducted amongst scholars, activists and governments, since their

appearance in the historical record and almost certainly long before that. The answer to the question

"Who are we? " inevitably includes ideas about "Who were we? " and importantly, "Where are we

from? " Most critically, it asks the question "Who are we not? " Modem historiography functions as

the history of nations, and frequently assumes that groups of people have who may or may not have

some shared characteristics have travelled through time and over sometimes enormous geographical

distances as sealed units. Each group is conceived of as existing in a kind of container, acting as a

barrier to keep the contents ̀ intact' (or `pure', as nationalists wish to present it), until their historical

trajectory or pre-destined "path" brings them into contact with history again, as a "nation". Those

that have lived through the experience of domination, especially by an "alien" culture and society

(usually an Islamic one, as in mediaeval Spain, or the Ottoman Empire for example), are frequently

seen as "... for centuries a people without history" (Bosworth, 1993: 190). They await in a pristine

condition, the establishment of nation-states based upon the particular identity of the group (Serbia,

Albania, Greece, Sweden, Norway, etc. ). Those `outside' the territory of this group are expected to

demonstrate an irredentist nationalism, militating to be united with their fellow nationals in a

`greater... ' entity. The importance of claiming particular external groups as one's own is not lost

upon politicians and ideologues; the Indian government's `sponsorship' of the Gypsies took place in

the context of growing tension with Pakistan in the late twentieth-century.

The concept of "nationhood", of belonging to a particular territory or geography in some innate,

primal fashion is of course, central to the nation-state. Those who are not of this "land", have no

"relationship" with the soil, encompassed in the idea that one's ancestors are buried in it or have

died on it are deemed to be "other", "alien" and "immigrant". Essential to the idea of "blood and

soil" is the concept of "sedentary", being `fixed' in some meaningful way (domicile, ancestor burial, defence against "immigrants", a notion of a common experience with others in the same

territory). At its opposite is deemed to be mobility, movement either in a regular or perhaps seasonal

pattern, or in a constant search for new resources, often associated with pastoralism. The meeting

point between these two poles is the confrontation between sedentary and nomad, or the "steppe

and the sown", as Harold Peake and Herbert J. Fleure have described it (1928). The suggestion that

there is a progression from one to the other is an example of European "Orientalism" (Said, 1972),

the view of the `Other' as exotic, existing in another, non-western time "... petty, barbarous and

72


cruel", as David Lean's Lawrence of Arabia describes the Arabs (1961; see also Hall, 1989: 21-34;

Ascherson, 1996: 61-4). In the context of Gypsy histories (the differing experiences of the Rom,

Dom, Lom and Travellers), this idea of a nomadic lifestyle giving way to a sedentary one is

frequently associated with the agent of time; "... many Romani are again living a nomadic life and

have returned to their earlier culture", according to an Azeri Rom (Daniels, 2003). "Settlement" is

identified as a movement from an older (primitive? ) past into a more modem (civilised? ) present,

rather than a particular point along a continuum, a condition that groups and individuals may move

in or out of at various times. The history of nomadic peoples demonstrates in general that the latter

is the most frequent mode of existence, but the ideology of the nation-state would seek to deny this

and promote the concept of sedentarism as part of "modernism".

The discussions about who and who are not included in these nation-states are, of course of especial

interest to us here; inevitably Gypsy peoples have been identified as ̀ not belonging' in most nation-

states at some time. In this instance, the Gypsies are usually seen as "... a people without history... "

eternally suffering persecution, discrimination, forced settlement or forced migration, though this

perception itself is recent. Before this, Gypsies were rarely seen at all, except as inherently

"criminal", but again without a history. This recognition of the experience of Gypsy peoples,

especially the Romani people is identified as happening in the moment; indeed a common

stereotype amongst non-Gypsies is the notion that these are a people who live in the eternal present,

incapable or unwilling to plan for the future or recall a more general past. Integral to the "scientific"

racism that underpins the ideology of the nation-state, the "People" as a "nation" exist on a point

along a continuum which stretches backwards and forwards in time (and often geographical space),

punctuated by events that are recorded and ordered in the national historiography. Other "nations"

or groups of nations may share a similar trajectory and are often perceived to further or less far

along it, depending upon their level of "modernisation", or what Samuel Huntington has

unashamedly described as "civilisation" (2002), suggesting irreconcilable differences between them

when identifying something called "Islam" and something called "the West". These differences are

most often presented in terms of "cultures", and hidden amongst them are the ideas of "progress",

"development" and "democracy" with their concomitants, "backwardness", "underdevelopment"

and "tyranny". Again, the idea of one "nation" or group of "nations" being more "developed" and "progressive" is contrasted with its opposite, to achieve the equation with "democratic" as opposed to "tyrannous". The Gypsies, especially the Romanies have been consistently defined vis-ä-vis this

series of Eurocentric ideas, even to the extent of characterising as "anarchical" political and social

73


organisation amongst some groups (though in fact, this prejudice extends event to groups that have

strong communal self-regulatory structures), in the "democracy-tyranny" spectrum. In this sense, it

is the concept of "origin" as a point of departure, which can be used as a "measure" against which

to compare others that makes the establishment of such more or less "important". The more

temporally remote that point, the greater the claim to an "authentic" longevity as a "nation" or

"people" and the more one's own "nation" may have "influenced" or informed others (Hellenic

Greece for example). Some or all of these considerations underlie the attempts to establish

definitively, the origin of the Gypsies.

In examining the record of references to the origins of the Gypsies (rather than records of the

appearance of Gypsies), inevitably one lights upon the most frequently cited example of both, the

reference to the "Lüri ' in the Shäh-näma or Book of Kings. The 60,000 verse Iranian epic was

written by Abu 1'Kasim Firdawsi. When completed, he presented it in 1010 CE to another important

figure in the discussion of Gypsy origins, Sultan Mahmüd of Ghaznä (997-1030). This `legend of

the Lüri and Bahrain Gür' has become ubiquitous in Romani scholarship, featuring in almost every

monograph, article and web-site devoted to the Gypsies and has been used to suggest their presence

in 5`h century Sassanid Persia. Recently critically re-examined and challenged on a number of points

(Hancock, 2003), the reference as first suggested by Col. John S. Harriott in 1830 attempted to draw

attention to the Oriental Origins of the Romanichal, or English Romanies. Firdaws and indeed his

earlier source Hamza al Isfähäni's Chronology, never defined their use of the term "Lüri 'beyond

explaining their presence in 11 ̀h century Persia by reference to a group of musicians from Sindh, or

western India in the 5'h century (see section 3.5). As there were many Hindus and Buddhists in

Sassanid Persia and the ruler Bahräm Gür (420-438) gave poets, musicians and singers "the highest

ranks at court" (Wiesehöfer, 2001: 159), little conclusive evidence that these were Gypsies can be

drawn. In the context of our current discussion, the suggestion that we might find the origins of the

Romanies in Persia as a group of Indians from Sindh must be seen to be a consequence of earlier

identifications coming, as it does in the 19`h century.

10th century Byzantine references to the Atsinganoi, originally a heretical Judeo-Christian group

centred in northern Asia Minor may be our earliest extant record of to Romani people, although

there are still questions about drawing equivalence between these references. The notion was

present in these that the group referred to were "... a Sarmatian people... " (Fraser, 1992: 46). The

Sarmatians were a Central Asian, nomad people, speakers of an Indo-Iranian language. Herodotus

in his Histories (Rawlinson, Rawlinson & Wilkinson, 1862) notes that they were a people who lived

74


and travelled in wagons, that their warriors included women and their priestesses men, and they

could retreat endlessly into their steppe lands, thus defeating the Persian Shähanshäh, Darius I

through despair and despond. Clearly, we are dealing with the Byzantine proclivity for using archaic

descriptions of contemporary peoples. Later, 12th century descriptions of Byzantine Gypsies are

much more reliable, referring to bear-keepers, magicians, soothsayers and snake charmers as

"Egyptians" in various religious commentaries and tracts (Soulis, 1961: 146-147). Numbers of other

references in late Byzantine sources indicate that there was no clear link made with the earlier

Indians in defining the origins of these wandering acrobats, jugglers and animal-trainers. One other

reference from this period suggests a connection with the Arabs, that of Simon Simeonis in 1323,

when he notes a group in the island of Crete who asserted "... themselves to be of the family of

Chaym... always wandering and fugitive... " and living in black tents similar to the Arabians' he

had seen elsewhere on his travels (Fraser, 1992: 50). The biblical reference to "Chaym" or Ham, is

frequent in the context of descriptions of the Gypsies in this period, especially in Western Europe

after 1400 when the Gypsies were identified as "pilgrims", atoning for apostasy and armed with

patents royal. Various were the legends that attached to these bands of distinctive, dark-skinned

travellers often led by `Counts' or `Dukes', but many of them made a connection with the fabled

land of Hermes Trimegistus, the `author' of a series of Gnostic and alchemical texts (actually

Arabic in origin; Holmyard, 1929: 525-6) believed to be `Egyptian'. Charters, such as the one

granted to Johannus Cinganus by the Venetians in 1244 at Nauplion, contained the evidence for the

later claims by Gypsy groups in Western Europe (Soulis, 1961: 164-165). In response to the

Ottoman incursions, Gypsy war-band leaders were awarded titles in charters such as ̀ duke' and

`count', with the concomitant responsibility for providing military service in the classic late feudal

relationship.

One of the earliest attempts to closely define the origins of the "Egyptians" in Western Europe was

by Andrew Borde (or Boorde) in 1547, when he published his examples collected in 1542, of "Egipt

speche". Survival of early works is often more by chance than by design, but some inference can be

drawn from the fact that original manuscripts are extant. Sebastian Munster's Chronographia

Universalis of 1550 also suggested an Egyptian origin, but in this case, Lesser Egypt, located as

Münster himself suggested, in the Gangetic or Indus regions (Bartlett, 1952: 85). Earlier, the

municipal authorities responsible for Hildesheim in Hesse recorded a visit from a party of `Tartars'

again from Egypt (Fraser, 1992: 66-67). `Little Egypt' was frequently cited as the place of origin in

various records of towns and cities in the 151h century, possibly gleaned from Gypsies themselves.

75


This was connected by commentators with the region of Modon, in the Venetian territories, but as

Fraser suggests it derived from the original notion of an `Egyptian' origin, as the community there

in the 14th and 15th centuries themselves claimed (1992: 53-54). One interesting reference to the

Gypsies of Modon at this time comes from Lionardo di Niccolo Friscobaldi in 1384, when he notes

that these apparent penitents without the city walls as Romiti (1818: 72). Other travellers to the city

suggested that it was the original home of this group, but this reflects the growing antagonism and

suspicion shown to Gypsies in Western Europe from the late 15th century, illustrating the shift from

pilgrims and penitents to `spies' and 'vagabonds'.

Their arrival in Western Europe, prompted by the chaos of the interregnum in the Ottoman lands

during the first decades of the 15th century and a worsening of conditions for the Gypsies there,

focused attention upon the `discovery' of the origins of this hitherto unrecorded group. The choice

by many annalists, commentators and writers of Egypt as the `home' of the Gypsies reflected the

concern with magic, conjuring and especially alchemy, following the 13th century translations of the

`works' of Hermes Trimegistus and their dissemination (Holmyard, 1923: 525). The appearance of

these dark-skinned, unusually dressed strangers also reinforced the ideas about the `East' as the

home of wandering tribes, heathens and infidels, the enemies of Christendom. The shift to the

biblical lands of Egypt may reflect the limited extent of the knowledge of those who first

encountered these groups in the Modon region and their assumption that these were people

connected with that land through magical practices. It may also be the case that, in order to allay

further suspicion, the Romitoi chose to present themselves not as quondam residents in the

Byzantine and then Ottoman lands (one a schismatic state and the other an infidel one), but as

people from a place that Europeans had some ideas of, the biblical lands. The flight to Egypt of the

Holy Family even provided some basis for claims to penitence; the Gypsies had refused succour to

the fleeing Christ-child and so were doomed to wander as a result. It is also possible that those who

gave Egypt as their origin, initially at least were indeed from that country. Their presence in Modon,

an important entrepöt for pilgrims to the Holy Land and other ports with traffic to the Latin

Crusader states like Ragusa (Dubrovnik), is compatible with such a suggestion. There is a

supposition that only Romanies (among the various Gypsy groups) were present in the Balkans, by

many modern scholars. The presence of an older community, absorbed into the more numerous later

identity and earlier instances of Dom being found in these regions, may be supported by the

reference from the Kingdom of Cyprus to the community of Gypsies there under the Lusignan King

Jacques 11 (1460-1473) (Marsh & Strand, 2003: 4). It is conceivable then, that the earliest

76


"Egyptians" spoke the truth.

Until the later 18th century, the `Egyptian' origin remained the dominant explanation, recorded in

tracts and treatises with little variation. The changing situation of persecution and oppression in

Western Europe required only that they be vilified by commentators as Ottoman spies, thieves,

idlers and con men and women. Ideological justification for the appalling treatment meted out to

Romanies in Europe was to be found in these works, and the earlier suspicions and prejudices took

on a lethal character when the writers of standing and influence took up their pens to do so. Even

the notion that Gypsies were indeed from Egypt came under scrutiny, and English encyclopaedias

referred to "... counterfeit kind of rogues, who being English, or Welsh people, disguise themselves

in uncouth habits... " according to Ephraim Chamber (1728). This terminology merely reflected the

earlier legislative descriptions of "counterfeit Egyptians" in a variety of punitive laws in the mid-

sixteenth century (Fraser, 131-137). Such an approach dominated the discussion of Gypsy origins,

and can be seen to manifest the beginnings of the racist paradigm of the "true Egyptian", as opposed

to the vagabonds and thieves claiming to be Gypsies (Fraser, 1992: 92).

The Indian origin of the Gypsies might be said to have been ̀ discovered' by Münster in 1550 CE,

although the claim that he was told this by Gypsies is not entirely correct (Hancock, 2002: 2) as

referred to above. Others did not take up his suggestion however, and presumably, the Gypsies

questioned by Western Europeans maintained the idea of an Egyptian origin, as it fitted well with

preconceived notions. The forty articles in the Vienna Gazette (Weiner Anzeigen) of 1775-1776 CE,

written by an anonymous Hungarian author (Fraser, 1992: 190), seem to have been based upon the

suggestion from Istvan Vali in the 1760's. Pastor Vali had allegedly attended Leiden University to

study religion (there is no record of him there), and came across a group of Indian or Sinhalese

students, from whom he collected about 1,000 words. Comparing these with the language of the

Gypsy labourers on the family estate in Raab, he `discovered' their similarity (Hancock, 2002: 2;

Fraser, 1992: 193; Hancock, 1991). Jacob Bryant also collected material, apparently at a Windsor

fair in 1776, which he published later (1785: 387-394). The German scholar, Jacob Rüdiger

collected examples of Romani from a Gypsy woman in Halle, and compared it with a variety of Indian languages, noting the similarities especially with a dialect of Lahnda called "Multani" (Fraser 1992: 194).

The most important work of the late 18th century was H. M. G. Grellman's Dissertation on the Gipsies (1787, English edition), originally published in Leipzig in 1783. In addition to the usual material describing the Gypsies in terms of stereotypes and prejudices, Grellman synbooked the

77


earlier Weiner Anzeigen articles and the work of other scholars, arguing for a clear relationship

between Romani and Indo-Aryan languages that was most closely related to Gujurat. He also

posited a date of departure from the Indian subcontinent at the time of the incursions by Timur-i

lenk (Tamerlane) into the Delhi Sultanate in 1398 CE (2d edition, 1807). The new science of

comparative philology guaranteed the interest of scholars in Romani and the origins of the Gypsies

as an example of change and development in an Indo-Aryan related language. Augustus F. Pott, in

his monumental The Gypsies in Europe and Asia (Die Zigeuner in Europa und Asien, 1844) and his

work on the Gypsies of Turkey (1844: 321-335), drew on the material that had by then proliferated

about the Gypsies and their language, to write the primary scientific works on Romani. Another

scholar, Franz X. Miklosich in Vienna (1872-188 1) wrote two volumes of history of the migrations

of the Gypsies, based upon the philological evidence, and can be said to be the first to re-construct

(or construct? ) the `long march West' of the Romani peoples.

Subsequent works investigating the origins of the Gypsies, and especially the Romani people have

all followed Grellman, Pott and Miklosich to a greater or lesser extent, apart from a few who have

sought to redefine the debates in ways that have reflected the concerns with the "social

construction" of ethnic identities. Okely's challenging The Traveller-Gypsies, sought to deconstruct

the `myth' of the Indian origin in regard of English Gypsies and demonstrate that the `link' to India

was a product of European Orientalism (following Said's book), an attempt to exoticise a socially

excluded and marginalised group (1983). More controversially, Willems, Lucassen and Cottaar

(1998), have developed a critique of the `traditional' perspective of Romani Studies and brought

their own "socio-historical approach" to bear, arguing that the Gypsy identity is a product of

Grellman, Pott and others since who have taken a widely disparate series of groups, who may or

may not share a number of characteristics and constructed a composite called "Gypsies". Taking the

notion of the "imaginary Gypsy" even further, Mattijs van de Port has argued that "Gypsy-ness" is

an "instance of the Wild", an aspect of societal discontent expressed in a cultural form and a

concomitant to "civilization" (1998), in Serbia especially. The `Dutch' school can be said to include

the important work of Ger Duijzings (1997), about the "making of Egyptians" in the context of the

Balkan conflicts in the last decade. His notion that identities can be reconstructed as necessity and

extreme circumstance demands, is the kind of psycho-historical analysis that Justin McCarthy

suggested in his catalogue of the `ethnic cleansing' of Ottoman Muslims from the region, 1821-

1922 (1995), when individuals traumatised by war and loss sought refuge in the `safety' of a

marginal identity. Without being too reductive, the underlying book shared by these and other works

78


is that, like Americans and many other modem èthnic' identities, "Gypsies" can be made and are

not necessarily "born".

Recent scholarship has sought to address both the implicit racism of the earlier `folklore' studies,

intrinsic to many of the contributions to the early Journal of the Gypsy Lore Society (1881-) and the

challenges of post-modern ideas about èthnicity' and ̀ identity' as a socially negotiated

phenomenon. Attempts to establish the history of the Gypsy peoples on a complementary basis to

the existing linguistic evidence in the formative lands of the Ottoman Empire, rather than to

construct it as Miklosich and others have done has demonstrated the extent to which the Romanies

(and consequently other Gypsy groups) are a composite people, from a variety of origins and the

product of complex social, economic and political factors (Hancock, 2003 & 2002; Mischek, 2003;

Altinöz, 2003; Marushiakova & Popov, 2000; Marsh, 1999). The argument that they were `forged'

in the borderlands of Anatolia between the hammer of the Saldjük Turks and the anvil of the

Byzantine Empire in the 12th century is becoming more widely accepted, though not uncontested.

The attempts to define more accurately the point of departure to specific raids by the Sultän

Mahmüd of Ghaznä in 1018 CE (Courthiade, 2003), demonstrates the extent to which there is a

`Romani historiography', seeking to refine and expand the knowledge about Gypsy origins and

history and confront some of the more extreme and ultimately untenable `myth-making' that has

passed for scholarship until now.

The key factor to the existence of the Gypsy peoples (Rom, Dom and Lom, Irish Travellers,

Yenische, Resande and others), has been the ability to exploit particular economic niches in

sedentary society, whether as commercial nomads, horse-dealers, farmers, metal-workers, miners,

gun-powder makers, canon-founders and carpenters or a host of other occupations that Gypsies

have undertaken to provide themselves and their families with a living. The musicians and

metalworkers of the Dom peoples (Ghagar, Nawar and Halebi) of Egypt, Syria and the Middle East

have adapted to their environment as it has changed to meet the challenges they have faced in a way

that other groups have not been able to so successfully. The Roma of South-Eastern Europe have

attempted the same, when faced with the horrors of èthnic-cleansing' and persecution as a product

of the assertion of ethno-nationalist identities in much of the Balkans in the last decades, fleeing

when possible and adapting to new environments in Sweden, the United Kingdom, Germany and

France amongst other states. The `resident' Gypsies (Romanichals, Resande, Irish Travellers, Sinti,

etc. ) of these nation-states have sought to maintain their distinctiveness vis-a-vis these newer

groups of Roma, whilst building essential relationships through political and cultural organisations

79


to challenge the common problems of discrimination, poor resources and marginalisation of all

Gypsy communities wherever they may be. The position of those groups adopting an exclusivist

and separatist identity, which is premised upon the notion of a more definite Ìndian origin' and

therefore of being `more Romani' (Hancock, 2003: 2), is mirrored by those who seek to re-affirm a

`Swedish' or Ènglish' Gypsy identity as ̀ more' legitimate in the context of the nation-state. The

search for the `true Gypsy' begun in Wh century Western Europe, ultimately provided the

ideological justification for the genocidal policies of emergent nation-states over the next two

centuries, as it, did in the period of Nazi Germany, 1936-1945. Such notions remain part-and-parcel

of the discrimination towards Gypsies in most nation-states of Europe.

The scholarship seeking to establish the origins of the Gypsy peoples must, it seems to me, refrain

from `buying into' such racist paradigms by following the patterns of nationalist myth-making and

nation-building adopted after the advent of Romanticism in late 18th century Western Europe, and

exported with such lethal results elsewhere today. The widely diverse and complex origins of

Gypsies in all their variation should be positively acknowledged, held up as an example to

challenge the absurdly reductive notions of the `Swedish', the `Norwegian', `Danish' or the

Ènglish' and other nations. The "liberation" that Said identified is in the "un-housed, decentred,

exilic energies... whose incarnation is the migrant... " (1992). The particular genius for adaptation

and flexibility that has been essential to the survival of the Gypsies stands against the rigidity of the

notions of the nation-state, which appear archaic and much closer to the fearful, exclusivist

`civilised' Athenians contemplating the `barbarian' Scythians in 4th century BCE Hellas (Ascherson,

1996: 62), than the trans-national, multi-cultural, diverse community of Gypsies in the early 2l St

century European context, examining their opportunities. To seek to separate out, to identify with

àrchaic' notions of ethnicity and territoriality, through tracing particular origins for their own sake,

may be a dangerous irrelevance, ultimately denying Gypsies the complexity of their history, rather

than explaining it. One particular piece of scholarship demands special attention, in the discussion

of origins, as it plays a seminal role in so many works and references to Gypsy history and

migrations that it has become almost axiomatic - the legend of Bahr .m Gür and the Lüri.

80


2.2 "... the strumming of their silken bows... " the problems of the legend of Bahram Gür and the

Lüli

THE REFERENCE TO THE FIRDAWSI LEGEND is one that is frequently cited in Romani Studies texts,

histories of all Gypsies, articles and newspaper reports (see Lori, 2003, for examples in connexion

with the Dom; Marushiakova & Popov, 2001: 11; Kenrick, 2004: 21; Fraser, 1992: 34-5), yet almost

no context or explanation is given as to who Abu `1-Kasim Hasan b. `Ali of Tüs, or Firdawsi (c. 329

AH/940 CE-411 AH/1020 CE) was, why he wrote the Shah-näma or Shähnäme, "Book of Kings"

(c. 1010 CE; see Cl. Huart, H. Masse, V. L. Menage, 2003: 918a; Warner & Warner's 1905-1925

English translation, 2005), and in what historical circumstances it was produced. Hamza al -Isfahäni

b. al-Hasan, ibn Mu'addib (c. 280 AH/893 CE-360 AH/971 CE), in his Chronology (Ta'r kh sini

mul(ik al-ard wa'l-anbiyä') of c. 961 CE, is an earlier source for the Bahräm Gür legend, for those

attempting to construct a "narrative of journey" for the Romani peoples during their earliest history

(see for example, Marushiakova & Popov, 2001: 11-12). Other "characters" (such as the King

Shangül of Hindüstän) have been merely treated as parts in a shadow-play, without investigation of

whether these have any basis in historical fact; like Karagöz, the Turkish Gypsy puppet, introducing

himself to us as a diversion from our worldly travails, they form a "backdrop" for the story. The text

itself has been little analysed and is quoted (most often uncritically, though Fraser does not;

1992: 36-7) in many cases, yet as Hancock remarks "it is wrong" (2002: 5). Here I would like to

suggest why.

Effectively with this tale, the perceived connection with an Indian origin for Romanichals (English

Gypsies), and by extension all Romani people, seemed confirmed, and an early date of departure

apparently established by the appearance of the LUr or LUl in Persia at the time of Sdsänid Shah

Bahrain Ghür, (420-438 CE). Hamza al-Isfaahdn also seemed to report an earlier version of the same

episode in his Chronology, c. 960 CE. With the production of an English translation of the

Shähnämä in India (see Macan, 1829) and a paper by Harriott (1830: 518-58) in the Royal Asiatic

Society's Transactions series, this story was seized upon as an explanation and `welded' to the linguistic arguments surrounding Romani origins. These suggested that one original migration had

left the north-western Indian region at a relatively early date, before separating into the three

distinctive linguistic branches of Romani, Doman and Lomavren (Marushiakova & Popov, 2000: 5)

somewhere in the Persian lands. The most influential of proponents was John Sampson, "... the leading English language Romani scholar of the early twentieth century" (Hancock, 2002: 3), who

published his work on the dialect of Welsh Gypsies in 1926. Through discussions of this Romani

81


monogenesis theory in the pages of the Journal of the Gypsy Lore Society [Journal of the Gypsy

Lore Society], Sampson's work was disseminated, and almost immediately challenged by Sir Ralph

Turner in his Journal of the Gypsy Lore Society article on Romani and Indo-Aryan (1926: 145-89).

Turner argued that he remained unconvinced of a singular origin for both Domara and Romani

(Fraser, 1992: 21), as the linguistic ancestors of each were related to differing groups of Indian

dialects, not the same (Hancock, 2002: 2-6). Despite this criticism, and continuing challenges from

more recent scholarship regarding Persia and claims for such early origins, both the Firdawsi

`legend' and the monogenesis theory are still frequently cited in discussions of language and

Romani history (see Hancock, 2006: 5-7; Mayall, 2004: 119-25; Fraser, 1992: 20-22 for summaries).

Some of the implications of this debate between Sampson, Turner and others, were that it

effectively focussed on key differences; firstly that the origins of the Rom, Dom and Lom peoples

as one proto or ancestral population, or "... the conviction that all Gypsies, dispersed at all points

throughout the world, were originally from a single stock. " (Mayall, 2004: 119) Secondly, that these

groups stem from entirely separate and distinct ancestry, sharing similar historical circumstances

surrounding their emergence as Gypsies (Hancock, 2000: 11). To some extent, the polarisation of

the two positions with their supporters and adherents has characterised the field of Romani Studies

ever since, in that these positions have become coalesced around notions that we might broadly

define as an ethnicised, or socio-historical discourses of origins (see Mayall, 2004: 3). Here we

might discern a crucial contest in the study of the Gypsies, between those who are committed to a

view of Gypsies as a distinct and identifiable ethnic group, with a history coterminous with other

ethnic histories (Kenrick, 2004; Hancock, 1985, for examples), and those who would see the claims

to ethnic identity as an aspect of political mobilisation, but not adequately convincing in the context

of scholarship and research (Willems, 1996; Okely, 1983). In this context, the legend of Bahräm

Gar becomes more than merely an interesting anecdote from an early mediaeval Persian source that

may refer to an episode in Romani history; it attains the status of "evidence" of claims to this

coterminous history, the ethnicised discourse of origins.

The context of the debate is important to establish, as it is essential to our understanding of the

competing discourses and, more importantly the longevity of this `myth' and its role. It is my intention to critically examine the principle elements of this oft-repeated legend; Hamza al-

Isfanhani's extraordinary Chronology of pre-Islamic and Islamic dynasties of Persia; Firdawsi's epic

of the struggle between good and evil, precipitated by murder and perpetuated through a bloody

cycle of revenge between the sons of Tur (nomadic Turanians from Central Asia) and those of Traj

82


(the sedentary Iranians). In addition, it is important to examine the processes whereby these

elements came to play a seminal role in the development of Gypsylorism and later, Romani Studies,

and examine the translations and references that were, and continue to be authorites in the

discussion of Gypsy origins. Finally, it is critical to decipher the character of the Säsänid Shah,

Bahram Gür (Vahräm V, 420-438 CE) in these works, before referring to Amir, later Sultan

Mahmüd of Ghaznä (389 AH/999 CE - 421 AH/1030 CE), the archetypal ghäzi ruler of his age and

emulator of much that is described in the cycle of legends about Barhräm, if we are to attempt to

understand the intentions of the authors of these episodes on their own terms, and see it as part of

the later writing about Mahmüd (Hardy, 1960: 25,170).

I suggest in this that we have a series of narratives recording the 5th century arrival in Säsänid Persia

of a contingent of allied Sindi mercenaries of Radjput origin, the remnants of which became

conflated with an eleventh century group of Domart itinerant singers, dancers and musicians, in

attempts to provide a plausible genealogy for the latter. This group was part of the wider community

of Gypsies that came to include elements from the later forced migrations of Sultan Mahmüd, those

known by the epithet of Käoli (now Kawli or "from Kabuli ', i. e. the central Ghaznävid territories),

and the descendants of an earlier Zutt population, especially from the ancient Indian colony at al-

Lür (Minorsky, 1986: 817b). The varied and differentiated character of the Gypsy communities of

modern Iran are, I would suggest, an outcome of this picture of complex origins, and the continuing

policy of forced population movements by the Khalifa and through to the late Ottoman state

(Windfuhr, 2003: 416a). It is also the case that the processes of the emergence of Gypsy identities in

Persia can be described in a way that mirrors the equally differentiated and complex picture found

in Europe. In the context of the semi-mythical chronicles and poetic epics of early mediaeval Iran

however, I suggest the tale of the Shah and the Gypsies must be seen as unreliable evidence of the

early arrival of any ancestral migrations of proto-Gypsy populations.

The chronicler and philologist known as Hamza al -Isfahäni b. al-Hasan, ibn Mu'addib (c. 280

AH/893 CE-360 AH/971 CE), was an accomplished scholar. He was especially known for the

meticulous lexicographical study of misspellings caused by the ambiguities of the Arabic script in

Persian literature, a study of Persian festivals, an extensively annotated diwdn of the most famous

poet of the `Abbäsid period, Abü Nuwäs al-Hasan b. Häni' al-Hakami (130 AH/747 CE-198

AH/813 CE), a collection of the proverbs and expressions of Persia, a work concerning the

superstitious beliefs and amulets amongst common people in Iran, and a political and biographical

history of Isfahän (Ar. Isbahdn). His greatest work was the remarkable Chronology, detailing the

83


history of the Islamic and pre-Islamic dynasties of Persia, and his survey of world history has been

studied in western Europe since the 18`h century and often translated since (Gottwald, 1844-48). It

would appear that although Hamza al-Isfahani was acutely aware of his position as a Persian man of

letters, and maintained some prejudices towards the Arab conquerors of Persia, he nevertheless

combined a thorough and original scholarship with a critical use of the best available sources,

whatever their provenance. His work "... demonstrates the breadth of enquiry amongst Islamic

scholars and the curiosity at work in Muslim scholarship in tenth century Persia" (Rosenthal, 1984:

156a).

His reference to the legend of the al-Zutt comes in his description of the life of the monarch,

Varakhän V (420-438), known to us as Bahräm Gür, or the "wild ass" (onager). Bahräm Gür V was

brought up by the Arabs at al-Hira and his education entrusted to the renowned scholar and jurist,

al-Mundhir Ibn al-Numän. Al-Tabari (c. 855) tells us that, contrary to legend, his appellation of

"GUr" was not achieved by a prodigious feat of archery in pinning a wild ass to a lion with one

arrow. Instead we learn that Bahräm, through constant exercise and physical skill was given the

name. Noted as a hunter by both his contemporaries and later by `Umar Khayyäm, it was eventually

tobe the cause of his death after eighteen years of rule (438). In his first year of kingship, he

declared war on the Roman Empire during the reign of Theodosius II, who had himself just

extended the city of Constantine through the construction of the Great Land Walls in 417. Despite

the Säsänid Persian success of winning Nisibin from the Romans, the general Mihr-Narse sued for

peace in 421 (Huart, 1983: 585b). In his own domains and those he conquered, Bahräm Gür showed

a propensity for the persecution of Christians. In the struggle with the Hephthalite dominated

Kidarite confederation, the Persian monarch personally slew the Hephthalite king at the Battle of

Kusmehan outside Marw al-Shähidjän in the contested territories of southern Trans Oxiana (Huart,

1983: 582b).

A number of stories regarding this monarch are given in Hamza's work, including one relating to

the "Treasury of Jamshid", a much-celebrated ruler of ancient Iran whose wealth Bahräm discovers

whilst out hunting and distributes to the poor, thereby enhancing his rule through this act of

kindness. This legend has its origins in the Shah's policy of tax remissions that he carried out at

points during his reign (Masse, 2003: 939a). Hamza's Chronology is partly in that tradition of

nahisal al-mulük or "mirrors for princes", a strong feature of Persian elite culture and an element of

statecraft in later Islamic imperial systems, such as the Ottomans (see Bosworth, 2003: 984b-988b).

The legend which most concerned Gypsylorists, and scholars of Romani Studies, occurs a little later

84


in the text, where he describes the story of the origins of the al-Zutt from the 12,000 Indian

musicians, sent by the King of India for the entertainment of Bahräm's bibulous, but poor subjects.

The story serves as the model for Firdawsi's later tale, and follows the familiar Persian pattern of

beneficence on the part of the monarch and the thriftless Zutt (Fraser, 1994: 34), as a `foil' for

Bahräm's virtues. The wide use made of the works of Hamza by later Islamic scholars doesn't

detract from the fact that there are some problems with his work. His lexicography suggests highly

unlikely etymologies for Persian words rendered ambiguous in Arabic script, revealing a proclivity

for invention and a bias towards looking for `evidence' to support his contentions about the

superiority of Persian, over Arabic (Bosworth, 2003: 985b). Additionally, his claims that the 12,000

al-Zutt dispersed into the Persian lands and multiplied, would seem to be contradicted by his

assertion that their contemporary numbers were small, yet he offers no explanation for this

disparity. Nonetheless, the use of Jewish, Greek and Armenian informants for sections of his

histories reveals a striking comparison with other examples of panegyric courtly composition and a

concern with veracity that others noticeably lacked (Robinson, 2003: 76).

Abu `1-Kasim Hasan b. `Ali Firdawsi was born 941 CE at Bazh in the Tabaran area of Tüs, to a

family of dikhans, or landowners in the village. He died in 1025-26 CE/416 AH (Browne, 1902-24:

ii, §2: 90). Like Hamza, Firdawsi was a passionate Iranian with a profound knowledge of the early

legends, myths and histories of Persia, gleaned from both Arabic and Persian sources. Some of

these became incorporated into the 60,000 verse epic Shähnämä, and again like Hamza, Firdawsi

made use of a wide variety of sources in producing his "Book of Kings'. He also extracted portions

from the work of his compatriot, Dakika, who had been assassinated by a Turkish slave sometime in

370 AH/980 CE, after which Firdawsl had begun to compose the Shähnämä. Dakika's rendering of

"an ancient book" that he refers to in his introduction, no doubt provided an initial inspiration; until

this point Firdaws had been the composer of some lyric verse and short, epic passages (Menage,

2003, II: 918a). Despite his historical association with Mahmüd of Ghaznä, Firdawsi only

approached the rulers of his day when he had exhausted his own resources and cannot be counted

amongst the other panegyrists, poets and historians brought to embellish and celebrate the court of

the Sultan, frequently against their wishes (Browne, 1902-24, ii §2: 91). Firdawsi's achievement

despite Mahmüd's miserly response to it (he gave only a very small sum to the poet, whereupon Firdawsi quit the court; Huart, Masse, Menage, 2003, ii: 917a), only serves to illustrate the

transcendence of the epic over much of the other panegyric output of the period (Huart, Masse,

Menage, 2003, ii: 918a).

85


Another important distinction was in the matter of faith; Firdawsi was of the Shi'i branch of Islam

whilst Mahmüd was allegedly militantly Sunni (Thapar, 2000: 29-31). Having secured the protection

and sponsorship of Mahmüd's first vizier, himself of Shi'i persuasion (an interesting indication that

Mahmüd was not the Sunni militant he is portrayed as), Abu `1-Abbas Fadl b. Ahmad al-Isfarayini

(994-1010 CE), Firdaws1 set about revising and extending his work, especially those passages

where he expressed his praise of Mahmüd, after the description of the death of Rüstara, for example

(see Warner and Warner, v, pt. Ii: 118; i, § 12: 112);

Abü'l Käsim! Our great Shah's hand is still

Thus generous alike to good and ill.

He never slackeneth in bounteousness,

And never resteth on the day of stress,

Delivereth battle when the times demand,

And taketh heads of monarchs in his hand,

But largesseth the humble with his spoils,

And maketh no account of his own toils.

Oh! May Mahmüd still rule the world, still be

The source of bounty and of equity!

As we might deduce, the amir was busy securing his reputation as Yamin al-Dawla `defender of the

faith', and Amin al-Milla `protector of the umma', and a prince on a par with Rüstara or Bahräm

himself, but with the fall of the vizier Abu `1-Abbas, Mahmüd's apparent intolerance for heterodoxy

became more pronounced (Huart, Masse, Menage, 2003, ii: 919b). The infamous reward that Firdawsi received upon submitting his magnum opus was clearly a comment upon this somewhat

opportunist change in opinion on Mahmüd's part. That it was opportunist is without doubt; the

support of heterodox, sometimes shamanist Central Asian elements in the Khorasan region where Iranians were predominant, was crucial to Mahmüd's early military successes in his expansionist

programme (Bosworth, 2003, VI: 65b). His role and the pre-eminent ghazi warrior was always tempered by pragmatism, and his maintenance of his Hindü troops, especially when deployed

against rebellious Muslim subjects, indicates that this ideology was part and parcel of the Ghaznävid ruler's `self-fashioning' (Hardy, 1960: 28). Firdawsi may have expected a more tolerant

86


and generous reception, if he understood the role of the poets and authors at Mahmild's court as part

of this process of promulgation of myth and majesty, and so his disappointment is understandable.

The primary problem concerned with both poets' work has been defined by most scholars as a

question of origins (see Fraser, 1992: 11-32 for a discussion of the various approaches this has

taken). Central to this problem and its exegesis, has been the endeavour to establish a coeval time-

line, matching the conclusions of those researchers for whom the analysis of the Romani languages

has provided the necessary ̀framework' for developing the history of the Gypsies (see Sampson,

1926; Turner, 1926; Gjerdman and Ljungberg, 1963; Kochanowski, 1979; for examples).

Frequently this has been at the expense of clearly establishing the relationship between language

and memory, as recorded history. The pursuit of evidence relies upon commonly assumed

connexions, as when Marushiakova and Popov refer to the Firdawsi episode as "... the events

described, although told in a semi-legendary fashion, and in much later times, are rooted in

historical fact and can be taken to refer to one of the initial stages of Gypsy migration" (2001: 11).

The main criticism of such presentations of "historical fact" might be summarised as follows;

Statements of this kind, even when they are partially true, ignore the principle that in order

to establish an historical connexion between A and B it is not enough to bring forward

evidence of their likeness to one another, without showing at the same time that the actual

relation of B to A was such to render the assumed filiations possible, and that the possible

hypobook fits in with all the ascertained facts... (Nicholson, 1914: 8-9)

Fraser (1992: 42) clearly cautions against reliance upon the single factor of language to determine

history, when he writes "... it is prudent to take stock of possible oversimplifications which [sic] the

linguistic approach to prehistory [i. e. early Romani history]... may encourage. " As such, the

lexicostatistical endeavour has resulted in a number of debates and disputes, assertions and

arguments, based upon abstracted notions of Romani history and migration in general, which have

been adduced from linguistics. In this context, reference to historical sources has often been

selective, and subjectively driven by the predisposition to support particular narratives. Uncritical

use of sources in some instances has led to misidentification of Romani peoples as other groups; a

case in point being the equivalence drawn between early Byzantine references to Atsinganoi or

Athinganoi and the Gypsies, despite Byzantine chroniclers' detailed knowledge of individual

heretical groups and their beliefs (see Hamilton, 2001: 72; Theophanes Omologetes, the Confessor

[c. 760-817 CE/c. 6,269-6,326 YC], cited in Hamilton & Hamilton, 1998: 36). Once again, Fraser's

87


scepticism proves salutary,

Too often the assumption has been made, in looking for traces of the Gypsies, that any

reference to a migrant group pursuing a Gypsy-like occupation can for that reason be

equated with them... (1992: 35)

In this instance, we might refer to the hostile reference from Theophanes to the Atsinganoi or

Athinganoi, (literally meaning "untouchables"), when he described the basileus Nikephoros I Genik

(802-811 CE/6,311-6,320 YC), requesting them to sacrifice a bull to subdue a riot (possibly with the

participation of certain Pavlikiani, or dualist Paulicians). Michael I Rangrave (811-813 CE/6,320-

6,322 YC) persecuted the Athinganoi severely, imposing the death penalty upon them for their

heresies and almost certainly reducing their numbers drastically (Mangrove and Scott, 1997: 739).

The distinct possibility that one of the Byzantine emperors was of this persuasion (Michael II the

Armorion 820-839 CE/6,329-6,348 YC), meant that conditions improved for this group and

Theophanes' continuator (writing about the period 813-961 CE/6,320-6,470 YC), suggests that this

was as a result of the fact that when growing up in Phrygia, Michael, a cattle herder at the time, had

been told that one day he would become basileus (Marushiakova & Popov, 2001: 14-15). By the

end of the 10th century, there are few indications that the Athinganoi were still in existence. The

argument that the term Athinganoi can always be taken to denote Gypsies in the Byzantine sources,

is inconsistent with the linguistic data concerning both Domart and Romani, that Matras suggests

cannot be accepted to have emerged prior to the 9th century CE at the earliest (2005: 68). The

descriptions of Theophanes Omologetes and his Continuator suggests a clear knowledge of the

practices of this heretical group (a mixture of Judaic, Zoroastrian and Christian), with no indication

of a differing language being spoken amongst the group, or an origin outside the imperial lands

(Phrygia and Likaonia). Speck's suggestion that this argument may be taken as given is, I would

argue, an instance of the impact of Gypsylorism upon Byzantine scholarship (as is the reference to

Athinganoi as equivalent to Gypsies, in Fögen's discussion of Balsamon on magic; see 1995: 102,

n. 15), but in this instance, it is not tenable (see Marushiakova & Popov, 2001: 14, for their use of

Speck to support this argument; Speck, 2003: 48). The more obvious references to Athinganoi in

Byzantine sources as Gypsies (those that are concomitant with the linguistic evidence), when one of

the grounds for conflation may result from the practice of Levitical ritual purity ablutions

(Bowman, 1991: 223) amongst the Athinganoi, something that might indicate the Aiguptoi

(Egyptians) maintained what appeared to be similar pollution taboos, or enough alike to be equated

with the Athinganoi. The practice of sorcery and divination is another.

88


In this current discussion, the case of the Lnri, Lors or Lori who are described in the Persian sources

have been firmly located in this discourse of origins, despite the problems of identifying who is

exactly meant by this description. As evidence of an early departure from India for the Roma, they

have been mobilised to support arguments between scholars, which have become extremely well

worn in the discipline through repetition. Indeed, it is almost axiomatic that the legend of Bahram

Gür and the Lür must appear in the early stages of any "history" or description of the Gypsies (see

Simmons, 2000, for an example from the "Origins of Gypsy Fiddling"). Many of these accept the

basic story as representing a factual, albeit couched in legendary terms, account of the earliest

migration (Simmons almost uniquely notes, "... modern scholars dismiss this story as romantic

fiction"). Such wide circulation has this particular episode had, that English folk-singers like Fred

Brookes can write a song about the subject and unquestioningly present it as part of the Romani

"tradition". The historical veracity of the story, the analysis of the descriptions Hamza al-Isf-ahäni

and Firdawsi (the two best known redactions) in either symbolic or semiotic terms, the textual

analysis offering wider perspectives and a more nuanced understanding of the descriptions, have not been widely undertaken by scholars. Despite the previous interpretations of this episode, and if

taken at face value, the story of a group of musicians from north-western India transplanted to

Persia in the mid-5th century CE remains just that. I would suggest that without further analysis it is

neither incontrovertible proof of a Romani presence in Säsänid Persia, nor is it yet a clear case of

mistaken identity, and thus the continuing uncritical use of this legend of Bahräm Gür and the Lüri

in any narrative of Gypsy history is indefensible.

89


2.3 The origins of the "Roman-i" connection

THE ORIGINS OF THIS LEGENDARY IDENTIFICATION are to be found in a piece written by a Colonel

John S. Harriott (sometimes misidentified as Captain James Harriott), of the East India Company

Army c. 1830. Colonel Harriott later became a Major-General of Her Majesty's Army in India

(1838) and was a Fellow of the Royal Asiatic Society, to whom he had submitted his treatise,

Observations on the Oriental Origins of the Romanichal, as part of their Transactions for that year

(Harriott, 1830,2: 518-558). Harriott was the kind of soldier-scholar familiar in both this milieu and

period, similar in many ways to the more famous Richard Burton, also an East India Company

officer during these years. His treatise closely followed upon a translation of the Persian epic

Shähnämä in four volumes by Turner Macan, published in Kalkhata (Calcutta), with the majestic

title The Shah Nameh... carefully collated with a number of the oldest and best manuscripts and

illustrated by a copious glossary of obsolete words and obscure idioms... that included a life of the

author in Persian and English (1829). Other European translations of Firdawsi's poem followed

this, indicating an especial level of interest in Persian literature by western European scholars at this

time. A French translation by M. Jules Mohl in seven volumes (Le Livre des Rois, Paris, 183 8-78),

an Italian verse translation published in Turin, by Pizzi (1886-88), German (F. Ruckert, 1890-05),

English (A. G. Warner & E. Warner, 1905-12) and Gudjarati versions (J. J. Modi, 1897-04) were

subsequently produced, to say nothing of selections in Danish (A. Christensen, 1931), Dutch,

Turkish and Özbek (Menage, 2003: 918a). The reasons for this rapid development in translations of

Firdawsi might be seen in a number of factors to do with European, especially British influence in

the region, as this was becoming dominant and the Empire strengthened control over the Indian

sub-continent, its resources and especially its trade. According to geopolitical logic, parts of the

"Middle East" were indispensable to the defence of this acquisition, in that the imperial mission

was seen to be justified by the earlier Muslim invasions of Firdawsi s patron, Sultan Mahmüd of

Ghaznä, as Sir Henry Miers Elliott suggested (1867-77: 3).

Thus, the work of Harriott, and others like Burton, must be seen in the complex light of European

Orientalism, and part of the process Said has described as

... to dignify all the knowledge collected during colonial occupation with the title "contribution to modern learning" when the natives had neither been consulted nor treated as

anything except pretexts for a text whose usefulness was not to the natives... (1987: 80)

Crucially for 19`" century European and Ottoman Orientalists, the article by Harriott suggested the

90


possibility of being able to "institute new areas of specialisation; to establish new disciplines; to

divide, deploy, schematize, tabulate, index, and record... every observable detail... " (Said, 1987:

86; Makdisi, 2002: 1-2), about an Oriental population in Europe's heart, the Gypsies. The

appearance of a group of itinerant musicians and thieves in Firdawsi's great opus, confirmed (for

Harriott and his readers) that the connection of the English Gypsies and the Indian origin of their

language could be made securely. This confirmation underpinned the founding of Gyspylorism, as a

new discipline and area of specialisation, a means of categorising "natives" in the colonies and at

home and of conceptualising the other in both settings. It is no coincidence that the investigations of

Harriott, Burton, and the later Gypsylorists are primarily intended to extend this categorisation, this

"mapping" of the Gypsies in their various "habitats". As Hancock (2004) has written, in his

introduction to the life and work of Jan Yoors,

... the same colonialism and the European domination of non-western peoples were feeding

into notions of a hierarchy of human groups. From the new sciences of botany and zoology

the move to classifying human populations was a natural step, and the idea of "races" and

their ranking occupied much of the scientific and nationalistic thought of the day.

Populations resulting from unions of different "races" were believed to inherit the worst

characteristics from each, and thus only the genetically pristine or "True Romany" counted for anything.

Whilst the work of Yoors was, Hancock argues, markedly different (2004), Harriott's study was intended to demonstrate the inheritance of genuine Gypsies, and those that followed him continued

to promote this true/false dichotomy.

91


2.4 The interpolations of Colonel Harriott

THIS TROPE OF AUTHENTtc/NAUTHENTIC followed upon both the much earlier deduction of Romani

as an Indic language, by Istvän Väli, Jacob Bryant, Jacob Rüdiger, and Heinrich Grellman, in the

late 18tß' century (see Hancock, 1993; Fraser, 1992: 191-197), and the notion of the "counterfeit

Egyptians" (Fraser, 1990: 43-69) that had been present since the mid-16t" century (Fraser, 1992: 85-

86). The migration to Persia in the 5th century appeared plausible, as it was alluded to in the

Shähnämä. The story suggests that Bahrain Ghür was visited by his Indian "father-in-law",

Maharaja Rao Shankal of Sind, who offered to send him 10,000 Lürl musicians to entertain the

ordinary Persians who were imbibing their wine without musical entertainment (Hancock, 2004:

para 12; Marushiakova & Popov, 2000: 5; Fraser, 1994: 35); although Hamza al-Isfahäni states the

figure of 12,000 (see Fraser, 1992: 35), whilst others suggest 4,000 in number (see Minorsky, 1986:

816b). The king was however, displeased with these Lüri and dismissed them before the year was

out. A number of inaccuracies have crept into the story, so that the most recent recapitulations of it

have conflated and reversed some important details. The 1905-25 translation by Warner and Warner,

Volume VII, Chapter 39 (148-150) refers to the episode in the following way;

Thereafter he sent letters to each archmage,

Gave clothing to the mendicants, and asked:

In all the realm what folk are free from toil,

And who are mendicants and destitute?

Tell me how things are in the world, and lead

My heart upon the pathway toward the light. "

An answer came from all the archimages,

From all the nobles, and the men of lore: -

The face of earth appeareth prosperous,

Continuous blessings are in every part,

Save that the poor complain against the ills

Of fortune and the Shah. ̀ The rich, ' they say,

Wear wreaths of roses in their drinking-bouts,

92


And quaff to minstrelsy, but as for us

They do not reckon us as men at all.

The empty-handed drinketh with no rose

Or harp. ' The king of kings should look to it. "

The Shah laughed heartily at this report,

And sent a camel-post to king Shangul

To say thus: "0 thou monarch good at need!

Select ten thousand of the Gipsy-tribe,

Both male and female, skilful on the harp,

And send them to me. I may gain mine end

Through that notorious folk. "

Now when the letter

Came to Shangul he raised his head in pride

O'er Saturn's orbit and made choice of Gipsies,

As bidden by the Shah who, when they came,

Accorded them an audience and gave each

An ox and ass, for he proposed to make

The Gipsies husbandmen, while his officials

Gave them a thousand asses' loads of wheat,

That they might have the ox and ass for work,

Employ the wheat as seed for raising crops,

And should besides make music for the poor,

And render them the service free of cost.

The Gipsies went and ate the wheat and oxen,

93


Then at a year's end came with pallid cheeks.

The Shah said: "Was it not your task to plough,

To sow, and reap? Your asses yet remain,

So load them up, prepare your harps, and stretch

The silken chords. "

And so the Gipsies now,

According to Bahräm's just ordinance,

Live by their wits; they have for company

The dog and wolf, and tramp unceasingly.

This text is the fullest English edition (available at http: //persian. packhum. org/persian/), and

generally considered to be the best critical edition, hence referring to it here. Harriott appears to

have translated the text himself in his essay of 1830, although he may have been using the four

volume 1829 Turner Macan edition, and this redaction differs markedly from the Warner in some

respects, most notably in Macan's translation of the apocryphal story of Firdawsi's fabulous reward

and extended sojourn at Mahmüd's court (Menage, 2003, II: 919b). The most obvious difference

between Harriott and the later version is the use by Warner and Warner of the terms Gipsy and

Gipsies, in place of his rendering of Lüri. The tale in both is more clearly defined in terms of

numbers and the change in conditions for these Lüri, their "contract" with the Shah. The translation

continues as above, until the final part where Harriott renders the text "... and support themselves by

means of their songs and the strumming of their silken bows... " (Harriott quoted in Fraser, 1992:

35, again a poetic detail absent from the Warner edition). Their dismissal also contains an

interesting difference, in that "... the Lin, agreeably to this mandate, now wander the world,

seeking employment... " and they are left to consort "... with dogs and wolves, and thieving on the

road by day and by night" (Fraser, 1992: 35), details not contained in the Warner translation. In this

instance, Harriott's insertion of ideas already associated with the concept of "Gipsie" are clearly

recognisable; the happy acceptance of their fate, as decreed by Bahräm, to wander, play and sing,

and the association of criminal activity with this perambulation. We can detect the ideas of the

author of the 1775-76 Wiener Anzeigen articles, and Heinrich Grellman's 1783 Die Zigeuner

(Fraser, 1992: 191-193) at work here, and Harriott's prejudices about Gypsies have been

94


interpolated in the text anachronistically, as the Warner translation suggests. The extent to which

Harriott is reflecting wider prejudices is also an interesting point; despite the use of the term Gipsies

by Warner and Warner, they do not seem to find the concomitant pejorative associations of petty

larceny in the Firdaws text. Clearly the Warner edition has been influenced by the widespread

acceptance, by the time of the publication of their series, of the tale as presenting us with something

about the origins of the Gypsies, so that the term in the Persian text has been equated with the

English term. In his introduction to Volume 7, Edmond Warner makes mention of the inclination of

"Professor Nöldeke... to consider Bahräm's importation of the Gipsies [sic. ] from Hind to Iran

historical" referring to Theodore Nöldeke's note in his Britannica article on the monarch and his

reign (Warner, 1905, vii, section 4: n). Again, this reflects the notion that Lüri can be equated with

the English term, Gypsy, but this does not prompt Warner and Warner to "gloss" the Firdawsi text in

the way that Harriott's earlier one does.

The other differences in terms of the Harriott translation and the Warner edition of Firdaws1 are

more significant, if less obvious in the former. The Indian ruler (Shangül; see above), is referred to

elsewhere in the text as the "noble chieftain of the Sindian host... " (Warner and Warner, 1905-25,

vii, §31: 125), and in a following section Rai or Rädjä (§37: 140), but the majority of the interaction

between the Shah and the King (§36-§38), takes place in Kanniij, as it is rendered (Warner and

Warner, 1905-25, vii, §29: 118), suggesting that the Gangetic basin is the heart of the King's

territory, that extends over the Sind. There is a long narrative of various fabulous deeds and exploits

on the part of Bahrain in Hind; he wrestles with the court champion after a feast (§28: 117), and

other feats, that precedes the reference to the Gypsies in the poem, in the tradition of the heroic

literature of the Khudäy-namag (The Book of Lords; c. 590-628). These deeds culminate in the King

of Hind offering one of his three daughters to Bahrain as a wife:

"0 thou Joy of hearts! thou hast prevailed.

Attempt no greater feat. I will bestow

My daughter on thee as thy wife, for thus

Shall I be profited in word and deed. (Warner and Warner, 1905-25, vii, §32: 127)

Herein lies the origin of the identification of the Shah as the son-in-law of Shangül; he is married at Kannüj to the "moon-faced maid" named Sapinüd, with whom he flees the intrigues of Shangül to

keep him in Hind, and returns to Trän (§34: 131-134). Reconciled to Shangül, he calls upon him for

the Lüri (§39: 148-150). It is in the consideration of this point that I will turn to a closer

95


examination of the Lüri, L511 and the Zutt.

96


2.5 al-Zutt and the Lüli

IF WE EXAMINE THE LITERATURE ASSOCIATED with this tale (Minorsky, 1986: 816b), it suggests that

the usage of the term lüri or LUl is itself inconsistent at an early point. Hamza al-Isfahäni refers to

the musicians in the story as al-Zutt in his Chronology (c. 350 AH/961 CE), but thereafter the terms

used by subsequent poets are related to Lüli, lüri, l6r ,

In the translation by M. Jules Mohl of

Firdaws (1838-78, vi: 76-77), the translator renders the term Lüriyän, and in his 1841 translation of

the Mudjmil al-tawar kh (c. 520 AH/I126 CE), Mohl extends this term to al-Ludy-an al-siidän, or

"the black Uri' (515,534). al-Tha`älibi writes in his Ghurar al-siyar or Ghurar akhbär mulük al-

furs wa-siyarihim (c. 429 AH/1037 CE), that the Lüri are descended from these "black" al-Lüriyün

al-südän (Zotenberg, 1900: 567), and following this, other Persian poets refer to the "blackness...

like night", of these people (there is no suggestion that they originate in the Sudan; see Minorsky,

1986: 817a). They are also described by writers as shükh "petulant", bunagäh (that is their way of

life is "irregular"), and most interesting of all in the context of the Shähnämä, shangül meaning

"extremely joyful" and "carefree in their happiness" (Minorsky oddly suggests this term means

"elegant" in his Encyclopaedia of Islam article; 2003, v: 817a). Modem connotations associated

with the term Lüli are similarly glossed (see Digard, 2003: 413b) though Hancock suggests an

etymology deriving from the Romanes with "lur" or "lot" meaning "robber" (Hancock, p. c. 2008),

whilst there are a significant number of terms associated with Persian Gypsies, both in terms of

occupational identity and regional designations (Minorsky, 1986: 817b; Digard, 2003: 412b-13a), to

which I shall return in the following section. This shift from al-Zutt to Lüli, al-Lünyan, al-Lfzriyan

al-südän, is not merely an alternative terminology, as demonstrated by the consistency with which

the latter term is used. It represents an alternative narrative, a differing interpretation from Hamza,

to the dominant discourse created after Firdawsi's text. I would suggest that Hamza is attempting to

include in his Chronology an historical account relating to an Indian population in Persia defined as

al-Zutt, one that is providing a plausible genealogy for groups defined as LUl or Lüri five centuries

later. Hamza is also attempting something else in his writing, for his text is one that is not

concerned with praising present rulers, but with detailing the Iranian past, in contradistinction to the

less glorious present.

Firdawsi's unequivocal praise for Mahmüd and his descriptions of Bahräm are intended as a

reflection of the characters of each, and an exemplar of the princely qualities embodied by both

monarchs. There is also the clear description of the lesser, feckless characters of the Lüri sent by the

King of Hind Shangül, almost certainly intended to pander to Mahmüd's own prejudices about

97


Hindus and their rulers. The cycle of events that leads to this episode demonstrates the duplicity of

the Indian princes through the characterisation of Shangül. The Hindu monarch is portrayed as (VII,

§29) deceitful, and cunning, intending upon bringing Bahräm to destruction by persuading him to

challenge a huge wolf (VII, §30) and then a terrible dragon (VII, §31). He even plots to have him

beheaded at his court, a deed so scurrilous that even his chief advisor will not countenance it (VII,

§32). Although the two are reconciled eventually in the tale after Bahräm agrees to marry Sapinüd

(though the couple flee to Iran), Firdawsi does not fail to point out the Indian remains "an idolater",

whilst Bahräm, he suggests, is "a worshipper of God" (although he presumably means Ahura

Mazda in this instance VII, §36), this is clearly intended to draw attention to the Shah's similarity to

the Sultan. In this (as in Bahräm's reply to Faghfür of Chin; VII, §35), the contrast is drawn with the

inferiority of the non-Persians, in their claims to majesty, their dealings with monarchs, and their

bravery and prowess. The argument could be made that Firdawsi was clearly appealing to Mahmüd

as a Persian monarch in the line of the Shdhanshäh or king of kings, and equally that Mahmüd

perceived himself to be so. Like earlier episodes in Iranian history, the Ghaznävids had secured

their position over their previous Samänid masters through these qualities, and thus had every claim

to be considered shahanshäh.

In this change in terminology as regards al-Zutt and the Uli, ambivalence arises that if uncovered,

may offer both the connection between the various ethnonyms, and provide an illuminating

perspective upon the origin of the Gypsy populations of Persia and elsewhere in the region.

Minorsky has identified in his article on the term LUli, that the origin of this name is in the early Arab scholars' description of the inhabitants of a town in Sind, called by them Arür or al-Rfir

(Alore; 1986: 817a). The Arab conquest of the region had taken this town sometime before 95

AH/714 CE, according to the historian al-Balädhuri c. 850 CE (Hitti and Murgotten, 1924: 439-

440). Muhammad Ma'sum "Nami" Mir records that Alore was

"... a very large city on the bank of the Mihran (the Indus); that there were many very fine

buildings in it; that outside and around the town there were gardens full of trees, having

good fruit, and that everything was to be found there that the inhabitants and travellers might desire"

and it was the royal residence of Rai Suheeris (Malet, 1855: 7). It fell to Muhammad b. 1'Kasim on "... Thursday, the 10th day of Rumzan, in the year Hijree 93 [711 CE]" (Malet, 1855: 17). The linguistic shift from Arbri/Rüfi to Löri/Lüli, Minorsky argues, occurs after the translation of Indian Alore into Arabic al-Mr, a dissimulation of the two "r" letters, being a common occurrence

98


(1986: 817a). The groups identified in the Shähnämä and other works, are seen as descendants of the

presumed captives from this, the most important city in Sind, after the Arab conquest in the

beginning of the 8t' century CE. This strongly suggests that the origins of the Dom are to be found

in such populations, a point I shall argue in the following chapter. What has happened in this

particular case is, I suggest, that the general term al-Zutt, the Arabic term for Djät, has given way to

the specific term Lüli, but that both have their origins in the same region (see Bazmee Ansari, 2003,

II: 488a). The semantic shift reflects a change in the presentation of the relationship between the

Sasänid shähs and the Gupta Kings of India, and the reconfiguration of relations in the wake of

Arab conquest.

The interpolation of the fabulous episodes relating to Bahräm Gür and Shangül King of Hind, in a

narrative depicting the prowess and bravery of the Shah, is a device to explain the alliance of the

Säsänids and the Gupta monarchs, in the face of a common enemy, the Hephthalite Huns, Hunas or

Hayätila (White Huns; see Enoki, 1955: 1-58; Howorth, 1873: 114-127). The origins of communities

of Indians as allied troops assisting the Persians in their defence, lies at the heart of the story of

Bahrain Gür, I suggest. The struggle against the Hephthalites was one waged by the shahs over two

centuries, from the initial attacks of the Chionite Huns in c. 350 CE against Shapur II. After a treaty

between these, the Huns refrained from full-scale assaults upon Iran until Bahräm Gür's reign,

though it is likely that sometime in the early 420's the Shah defeated the Hephthalite king,

dedicating his crown, the Hephthalite queen and her servants to the Gushnasp fire-temple at Shiz

(Morony, 2003, IX: 74b). Later onslaughts were not repelled, and the Hephthalites came to

dominate Persia and India, as I shall discuss in the next section. However, the role of victorious

allies that the Hindus had played in relation to the fifth century Sasänid shah became problematic

for the later, Muslim Persian chroniclers, as the Indians maintained their Hinduism, even after the

Arab conquest of Sind, thus remaining "idolaters", whilst the Sasänids could at least be represented

as believers monotheists (adherents of Ahura Mazda), and in some senses closer to Islam. The

legend of the L611 functions as a semiotic dislocation describing this shifting relationship, giving an

ignoble origin for people who were once valued and respected allies.

99


2.6 "Strong and valiant men"; the Hephthalites

THE HEPHTHALITES APPEAR IN ROMANI STUDIES literature as the impetus for the earliest migrations

during the 5th and 6th centuries CE (Marushiakova and Popov, 2001: 11-12), or more specifically for

the Dom in the context of groups of composite military units moving westwards to confront the so-

called "White Huns" as they swept into north western and central India (Hancock, 2000: 9-11). The

argument has been made that the Hephthalite Huns are a contributing `push' factor in the migrations

of the early Gypsies, and as such it is to them that I would like to turn in order to discuss the

question. The origins of the Hephthalites are very obscure and frequently contradictory theories as

to whether these people were Turks or Indo-Iranian have been put forward (Litvinsky, 1996: 135-

144; Sinor, 1990: 294-5,298-301; Frye and Sayili, 1943: 195-6; Enoki, 1955: 56). Enoki's argument

that the (H)Ephthalites should be regarded as an Iranian tribe have been cautiously accepted by

Frye and Sayili and others, though they suggest that like most central Asian nomadic empires, the

Hephthalites should be seen as a composite confederation containing differing elements including

both Turkic and Iranian (1943: 195). Sinor suggests that the term is a dynastic appellation that was

adopted by the Avars towards the close of the 50' century CE, when Hephthatlite rule was

consolidated over Tokharistan, where these people had been living for some time (1990: 298). This

new confederation pushed the earlier Kidarite rulers westwards into a confrontation with the

Parthian and Sassänid empires (1990: 299). The Hephthalites themselves were to expand into

Sassänid regions at the expense of the Persian rulers for a period, though earlier they had been

repulsed by Bahräm V GUr (420-38) as the Shäh-näma tells us (Warner and Warner, VII, pp. 165-

69; Litvinsky, 1996: 135), whilst also noting the names of the Hephthalite rulers as Iranian.

The term Hephthalite itself has a variety of meanings given in the sources, but most can be

associated with the notions of `bravery', `strength' and "valiant men" (Litvinsky, 1996: 135).

Chinese sources suggest these people are descended from the Great Yüeh-Chih mentioned in the lists of tribute bearers at the Chou court, 1000 BCE (Narain, 1990: 155) in the region of Tokharistan

and Gandhara and are known as I-ta (Enoki, 1959: 7). They are also mentioned as originating from

the Ch'e -shih of the Turfan oases, and numismatic evidence suggests the use of a Bactrian script and an elite language derived from East Iranian (Litvinsky, 1996: 135). The Chinese suggest that the Yüeh-Chih spoke a different language from all those around them (the Juan Juan, Kao-Ch'e and Hsuing-Nu, all of whom were Altaic Turkic speakers; see Enoki, 1959: 6), but the linguistic

evidence is slim indeed (Sinor, 1990: 300). These people may have been Indo-European speakers inhabiting the Tarim Basin and the Taklamakan Desert region (Narain, 1990: 154). In the 5`h century

100


BCE this area had been home to the trading peoples who appear to have been from Central Europe

and Celtic in origin, weavers of particularly fine cloths, with a burial culture preserved in the so-

called "mummies of Ürümchi" (Wayland Barber, 1999; Mallory & Mair, 2000) and the Hephthalites

are suggested to have originated in this area (Enoki, 1959: 3). Enormously successful traders along

the ancient silk routes, the peoples of the Tarim basin created a complex and flourishing culture

connected through trade to the Chinese Empire and those of the Middle East (Wayland Barber,

1999: 12).

Procopius of Caesarea noted that they were a "... Hunnish people and so called [but]... do not mix

and associate with those Huns we know... " (Procopius, 1966, quoted in Litvinsky, 1996: 136),

indicating that the Hephthalites were not Huns but used this appellation to intimidate their enemies

(Frye, 1984: 346). Described as settled, "white-skinned and not ugly or bestial like other Huns... "

they were ruled by one king and "... possess a legal state structure, observing justice amongst

themselves and their neighbours in no lesser measure than the Byzantines and Persians... "

(Procopius, 1966, quoted in Litvinsky, 1996: 135). Chinese sources however suggest that the

Hephthalites had no cities and lived as nomads, based upon eyewitness accounts (Sinor, 1990: 300)

thus illustrating the complexity of defining who exactly these people were. The question of

language is also problematic, as Litvinsky suggests (1996: 136) that only a very few words are

known from the numismatic evidence, but the most modem scholarship has argued that the nature

of the confederation was such that the elite may have spoken an Iranian language (as suggested by

the inscriptions upon these coins; Bivar, 2008) as it may illustrate and elucidate the complexities

noted by the sources between the so-called "Red" and "White" Hepthalite Huns, shown in mural art

of the period (Litvinsky, 1996: 137). These appellations may not be related to physical differences

however, as central Asian nomadic groups often had these definitions added to ethnonyms (Sinor,

1990: 301). These differing perceptions of Byzantine, Persian and Chinese sources is perhaps best

explained by the representations we see in the murals, depicting the remnants of separate socio-

economic and ethno-linguistic groups (Litvinsky, 1996: 141). The Hephthalites would appear to be a

complex confederation depicted in the sources and secondary literature in a variety of ways and

using a bewildering number of ethnonyms (Ephthalites, White Huns, Hayätila, Hunas, Xyön, Hyön,

Var, Avar) and whose origins and ethnological attributes are extremely unclear. The Hephthalite

state may have been based upon a nomadic, Indo-Iranian elite who governed a largely sedentary

society of primarily Turkic and Indo-European peoples, and was certainly powerful enough to

establish control over an enormous area from Tibet to Turkmenistan in modern terms (the borders of

101


the Northern Wei and Southern Qi dynasties of China to the eastern Sassänid Empire). If as Sinor

suggests the Hephthalites are to be identified as a dynastic appellation of the Avars (1990: 298,301),

then the inclusion of the lands of the Avar Khanate would suggest that the northern limits of the

state stretched from Siberia to central India. The state would appear to have lasted from the early 5th

century when it broke away from the Juan-juan confederacy in or around 415 CE (Sinor, 1990: 293)

until its destruction by the Türks between 557-561 CE (Sinor, 1990: 301), although some

Hephthalite elements continued to exist as late as 651 CE, when they fought against the Arabs in

Khoräsän, and in 704 CE as part of a rebellion in the same region (Bivar, 2008). It is with the

conquest of large parts of northern India that the Hephthalites have been connected to the ancestors

of the Gypsies (Marushiakova & Popov, 2001: 11-12), or at least the Dom (Hancock, 2000: 9-11), as

I suggested at the opening to this section, though the case has not been made in detail. In one other

instance the Hephthalites have also been cited in connection with sun-worship and Kosovan Roma

beliefs about the importance of the sun (Polansky, 2006: 67). The latter draws upon the detailed

descriptions in Enoki regarding the customs of the Hepthalites in this regard (1959: 45-9), and then

suggests that the city of Multan was, in the pre-Islamic period, a centre of sun-worship in India

(2006: 65). Other sources are less clear about the nature of the Hephthalite beliefs; Enoki quotes the

Chinese envoy Sung Yin in describing them as being antithetical to Buddhism (1959: 45) and

building their tents to face eastwards, indicating a belief in sun-worship (1959: 46), though possibly

also elements of Zoroastrian fire-worship. There may also have been practices linked with the

Turkish Tang'ri or Sky God, and elements from Hindü Shiva worship (Enoki, 1959: 48). Their

attitudes towards Buddhism did not preclude a tolerance shown towards Buddhist shrines and the

worship of their subjects in them (Enoki, 1959: 49), though Litvinsky suggests that in the conquest

of India (477-520 CE) the Hephthalites destroyed many hundreds of Budhist temples and stupa

(1996: 147), though this is another contradictory factor, given the tolerance of Buddhism in Balkh

according to Chinese eye-witnesses (Enoki, 1959: 47). A complex religious picture emerges, with

what seems to be a profoundly syncretic pattern of beliefs amongst the differing groups within the

Hephthalite state (Litvinsky, 1996: 142-4) and differing relationships between these groups and the

ruling elite over time. It may have been that the practices of central Asian nomadic invaders from

time immemorial were behind the initial destruction of the Buddhist places of worship, thus

terrorising the population and eliminating the possible resistance from groups in Indian society,

followed by a more conciliatory policy once the conquest was complete. The destructive and cruel

hostility of the Hephthalite ruler Mihirakula (520-42 CE) towards Buddhism during the initial

invasions, contrasts with the evidence from a seal that shows him worshipping Hindü deities Vishnu

102


and Shiva, inscribed in Bactrian script and dated to the late 5th century (the seal is currently in the

collection of the British Museum) so perhaps individual rulers adopted varying attitudes to the

religion of their conquered subjects.

The Hephthalite destruction of the Gupta Empire followed an earlier repulsion of the Kidarite

invaders by the Indian prince Skandagupta in 455 CE (on behalf of the Emperor Kumäragupta, d.

455 CE), almost certainly prompted by the pressure further eastwards on the Kidarites from the

Hephthalites (Bivar, 2008; Litvinsky, 1996: 141). The death of Skandagupta in 470 CE marked the

turning point for the Hephthalite invasions, as they had previously subdued the Kidarite

confederation and began embarking upon the conquest of India in earnest in order to control trade,

spices, tea, silk and slaves (Barfield, 1989: 2-4). The conquest of northern India and the subjugation

of the Gupta Empire lasted until the middle of the 6th century but would seem to have finally ended

with the victories of the Türk kaghanate and those of the Sassänids, by which time the Hephthalites

had already been repulsed from the region of northern India, though it is not clear if the dynasties

that took control in the regions of Gurjara represent some continuity with the Hephthalites, in ruling

Kannauj for example (Bivar, 2008).

The connection with the early migrations of the ancestors of the Gypsies is extremely tenuous as the

documentary evidence does not suggest large scale populations movements resulting from the

invasions of the Hephthalites and the scenario presented by Hancock, of composite armies

marching to confront the `White Huns' outside of the territories of the Indian lands is also at odds

with the evidence of Hephthalite incursions. The defeat of the Hephthalites by the first Mawla

prince Yasodharman in 528 CE (Malcolm, 1880: 191-2) was one that took place upon Indian soil,

not in the territories beyond the Hindü Khush and the final destruction of the Hephthalite state did

not involve the Hindu kingdoms that emerged from the break-up of the Gupta Empire. The

likelihood of the Hephthalites being the mechanism by which large numbers of Hindus migrated from India in the 5th century is very uncertain, particularly when a much clearer trajectory of forced

migration can be suggested with the Arab conquest of the Sind at a later period, as I shall argue

subsequently. The strength of these connections would seem to have relied upon an initial

identification with the Hephthalites in the Shäh-näma and the work of Col. John Staples Harriott,

and the arguments for accepting the equivalence of the Lüri with the ancestors of the Gypsies.

These are I would argue here not tenable, and therefore the suggestion that the Hephthalites are

related to the events that gave rise to the Romani people as an aspect of this can no longer be

accepted.

103


2.7 Harriott, history and historiography

IN THE CONTEXT OF THE ABOVE ARGUMENTS, the seminal role of John S. Harriott, in the

identification of the story of the Lüli and Bahräm Gür, with the origins of the Gypsies, must be re-

evaluated. Harriott's translation of this section of the Shähnämä. (1830: 518-58; Warner & Warner,

1905-25, VII, §39: 148-150), interpolated anachronistic notions relating to the character of Gypsies,

as these had been defined in European scholarship since the 1770's and had been part of the popular

prejudices and stereotypes for a great deal longer. Harriott's glossing of these notions upon the text

of Firdawsi's story, added an additional layer to an already complex text; one that contained

elements of the less subtle panegyrics being produced at the court of the sultan Mahmüd of Ghaznä

as a supplication in times of the author's needs, yet retained a transcendent narrative and structure

that lifted it beyond these material concerns. Firdawsi's religious heterodoxy may have added an

additional motive to those of Mahmüd in awarding the aged poet a meagre pension for his

monumental work, but the text itself displayed a clear intention as regards the comparisons of

Bahräm and Mahmüd in this tale and the tropes of the deceitful and dishonourable behaviour

reflected the prejudices of both poet and sultan towards the Hindus, I would suggest. Harriott's

colourful redaction of the text concerning the Lüli or LUr has fundamentally been at the base of a

positive identification for many scholars, with the Gypsies, yet this reference has not been

systematically or rigorously interrogated by either Gypsy-lorists, or modern Romani Studies

scholarship. Upon examination, the translation by Harriott displays a number of aspects that throw

doubt upon whether this may be safely asserted in any connection with Gypsy origins, and I would

suggest that the continuing use of the reference is an aspect of the mythologising of Romani history

that must be separated from the actuality of that history, whilst it may continue to be described as an

aspect of Romani historiography.

104


3.1 The question of an Arabic `lingua poetica', and the origins of Domart in the Umayyad Khalif a

"Lo! It is Allah who splits the seed and the date stone... He brings forth the living from the dead... " (The Glorious Qur'an, Al An 'Am Sura: 95)

IN THE LIGHT OF THE PRECEDING DISCUSSION regarding the origins of the Gypsies and the legend of

Bahräm Gür, it will be important to examine in the next chapter, the evidence for suggesting that we

may locate Domart origins in populations of Hindü captives from al-Sind or Sindhu (Sind,

Belücistän, Makrän, parts of the Pandjäb and Rädj Qästhän, as described by Arab geographers;

Bosworth, 1997: 632), and al-Hind (generally referring to the regions east of the Indus or Mihrän, in

this period; Maqbul Ahmad, 1986: 404-5). The regions were the subject of Arabic invasions and

conquest in the first century of the expansion of Islam, and in this context, it is necessary for me to

predicate my considerations with some discussion about the nature of pre-Muslim Arabic society, its

transmogrification through the impact of Islam, and the intrinsic relationship between Muslim

society, and the formation of what we might call proto-Domari groups in the Iraqi (Iräki) lands of

the later Khalif'a. In particular, I concentrate on the emergence of an Arabic koine, as a precedent

for the eventual emergence of Domari, and a military context for the coalescence of Dom identity in

the armies of the Umayyad Khalif'a, in the conquest period, during the first two centuries of Islam.

The most obvious location for this precedent would be the so-called poetic koine of the pre-Islamic

early 6th century CE, but for reasons I discuss below, I suggest this is not the case, as this `lingua

poetica' is itself uncertain, and contested by some scholars and the evidence for its early existence

debatable. The model for the emergence of the Domari-speaking community is to be found, I

suggest, in the Arabic armies of the conquest period, the first two centuries AH of Islam (7`h and 8t'

centuries CE). The re-configuration of Arabic society as a result of the revelation of Islam, the

removal of the Prophet (peace be upon him) and his followers to Yathrib or Yasrib (al-Madina after

the phrase madinat al-nabi, "the city of the Blessed Prophet"), in 622 CE/I AH, and the campaigns

of expansion and incorporation of differing populations carried out subsequently, provide a better

precedent as I will argue, for the evolution of Domari and the Dom.

The capture of many thousands of Hindus and their forced migration to the Umayyad heartlands is

the primary factor in a chain of events that results in the eventual emergence of the Dom and this I

suggest, prefigures later developments that lead to the ethnogenesis of the Romani people. There are

common factors here I propose, in the pattern of confrontation, defeat, capture, forced migration, a

reconfiguring of the social organisation of these Hindus with the collapse of their internal system of

social stratification, belief and value systems, that undermined the existing socio-cultural structures.

In addition to a linguistic coalescence under these conditions of a societal shift, new forms emerged

105


as a composite identity, reflecting the diversity of the groups involved in the process. This process is

one that is intrinsically bound up with the nature of Islamic society as suggested above, in that the

social organisation of the early Muslim state contained the possibility of incorporating and

reconfiguring disparate populations in new ways. This point, although perhaps controversial is one

that is at the heart of the processes of ethnogenesis of the Gypsy peoples and must be dealt with

primarily, as it underpins the conceptualisation of this process as argued throughout this book.

The traditional portrayal of this period of the early part of the life of the Blessed Prophet, accepted

by many scholars (though not uncritically; see Crone and Cook, 1977), suggests that Arabian

society was in the process of significant change and to some extent, dislocation (Lapidus, 1988: 11).

In the period just prior to the revelation of the Apostle of God and the advent of the community that

became the umma, Arabian society "was in [a]... ferment" (Lapidus, 1988: 20), in which "new

conceptions of collective identity emerged" (Lapidus, 1988: 18). Mecca (Makka) was the site of a

diverse population without tribal affiliation; refugees, merchants and foreigners, all of who held

various religious convictions and moral conventions (Hourani, 1991: 11). The shift from collective

to more individual notions of responsibility and social status was beginning to emerge here, and the

reconfiguring of social relationships based upon these ideas was starting to have an impact upon

Arabian society outside of the harm of Mecca. Lapidus attempts to encapsulate this society as one

that "set individuals free from the traditions of their clans and allowed for the flourishing of self-

conscious, critical spirits who might conceive of a universal God and universal ethics. " (1988: 20)

The strongly conflictual tendencies present in Meccan society at this time could have been

unleashed and eventually eradicated this social grouping, but instead its coalescence around the

Prophet Muhammad (pbuh) and his revelation, and the relocation of the community of his followers

to Yathrib (al-Madina) offered the opportunity to establish this emerging religious and social group

on a more substantial basis. The cohesion of this community was forged through common belief,

the development of a social and ethical system, alienation from the surrounding social structures

and the evolution of a language of prophecy as the foundation for the expression of this revelation

from God, vouchsafed to the Blessed Prophet. This discourse and the process of seeking to define it

in order to successfully transplant the divine message beyond the boundaries of the initial

collectivity, thence the ethnic group (Arabs) to achieve the universality made explicit in the

message of Islam, is the foundation for the emergence of a koine that underpins the varieties of Arabic that we know today.

One of the aspects of this "ferment" and one that is essential to these considerations, is the

106


discussion about the emergence of this koine. Established scholarship identifies this with a

particular form of common language, al-'Arabiyya, used throughout Arabia by reciters of epic

poems (djähiliyya), or ruwät [sing. räwi] (much like the Dom in present day Egypt), that "marked

the development of a collective identity transcending the individual clan". (Lapidus, 1995: 18)

Hourani remarks that there was

a growing sense of cultural identity among the pastoral tribesmen, shown in the emergence

of a common poetic language out of the dialects of Arabia. This was a formal language, with

refinements of grammar and vocabulary, which evolved gradually, perhaps by the

elaboration of one particular dialect, perhaps by a conflation of several... Their poetry may

have developed out of the use of rhythmic, elevated and rhymed language for incantations or

magical spells... (1991: 12)

The use of this kofne was, it is suggested by Hourani, common at tribal gatherings, in markets,

towns and oases, and also at the courts of the Arab dynasties located on the peripheries of the great

empires of Byzantium and Persia: the Lakhmids in Hira and the Ghassanids of Syria. It apparently

crossed social and cultural boundaries to become eventually, the medium of classical Arabic literary

expression, al-Mu'allaqat, or `suspended poems', from their supposed hanging, written in letters of

gold, on the Ka'ba, generally numbered at seven. This etymology is more likely, in fact, to refer to

the poems, the Shi'r or "Seven" as a "necklace" or hanging jewels (Lecomte, 2003, vii: 254a). The

process of the development of this poetic lingua franca, evolved out of a much older tradition over

time and was not recorded until later, as this was a form of expression that relied upon public

recitation (and still does, in the case in the case of the Dom reciters of the epic poets; see Reynolds,

1995). A discussion of al-Mu'allaqat also provides a context for examining the composition of the

Qur'dn as a relatively uncomplicated process, by implicitly suggesting the means of expression

already existed for the setting down of the revelations to the Blessed Prophet. The question about

the differing interpretations and the redactions of the text remains beyond the bounds of this work,

yet it must be noted that the process of setting down the actual text of the Qu'ran was one that

remained problematic until the introduction of a fully vocalised and pointed text, in the `Abbäsid

period, popular tradition ascribing this to the governor of Iraq (`Iräk), al-Hadjdjädj (74-95 AH/694-

714 CE), resolving some of the previous confusion resulting from a proliferation of interpretations

during the Umayyad Khalif'a (41-132 AH/661-750 CE; Paret, 2003, v: 408a). The implicit

argument made by Lapidus, Hourani and others is that the process of linguistic development had

been part and parcel of the emergence of the Muslim identity, a recognisable aspect of the evolution

107


of this group in Arabic society.

The form of Arabic that is identified by St. Jerome as arabica lingua in his Praefatio in Danielem

and in Hebraic sources, is what may be reflected in the term lisän al-'arab, the source of literary

Arabic (Rabin, 2003, i: 561b). The description al-'arabiyya al-fushä, can be understood as something

more closely related to the notion of an emerging koine in the sense of "clear" or "universally

intelligible", though not a synonym for "pure Arabic". The discussion of the development of Arabic

as a script, linked as it is with the evolution of literary Arabic, is also of interest in some specifics,

such as its invention by Christian missionaries at al-Hira (interestingly enough, an example of the

earliest Arabic script being found at the church of Hind, or the Christian Indian church from c. 560

CE), as a means of propagating the Bible amongst the Arab tribes (Abbott, 1939: 5 plate 1). Muslim

authorities suggest the passing on of written Arabic by two Christians from al-Hira, Zayd b. Hamad

(c. 500 CE) and his poet son, to the Arabs, in some form that was part way to the later classical

language, as it was described as "not clear" (Rabin, 2003, i: 565a). The development of the form is

one that seems to have been a process of evolution over some time, and it is recorded that the poet

`Adi b. Zayd b. Hamad drew upon many different dialects in his, and his father's work, suggesting a

complex trajectory with associations to notions of the superiority of some of these dialects (that of

the Quraysh, the Blessed Prophet Muhammad's tribe, for example) and to the Bedu, especially of

the Nadjd, who remained the locus of notions of "genuine" Arabic into the Islamic period (Hourani,

1991: 13).

The Qur' änic evidence would seem, at first glance, to support the notion of a poetic koine at an

early date, mutually comprehensible amongst the various dialects of the Arabian peninsula, as it is

identified by early Muslim scholars as the "clear Arabic Book" (Khalafallah, 2003, i: 565b-569a).

This is evidenced by the connection between the djähiliyya poetry (the Mu'allaqat), and the

structures, syntax and grammar of the Holy Qur'an. The process of assembling the authoritative

version of the Glorious Qur'an, apparently took place during the period of the third Khalif Uthmän

(644-56 CE), although Muslim tradition ascribes the "collection" of the Holy Qur' än to within a

short period of the death of the Prophet (peace be upon him), during the Khalif'a of Abü Bakr (632-

4 CE), but this tradition is problematic (see the comprehensive discussion in Welch, 2003, v: 400b-

427a). The complex process whereby the final, consonantal and pointed version of the Qur'an was

produced is difficult to identify, and the question of variations in the original "readings" or ways of

vocalising the texts, is one that relies upon the science of the hadiths or traditons regarding the

sayings of the Blessed Prophet, as a means of corroborating understandings of complex and

108


potentially ambiguous material (the "Satanic verses", for example; see Welch, 2003, v: 404a; All,

2003: 26-7). To this extent, it may not be possible to rely upon the Qur'änic canon as evidence for

the existence of an earlier, pre-Islamic koine "clear" and "universally understood", in this sense

(Rabin, 2003, i: 561a).

The theory that a pre-Islamic koine was in existence then, relies to a large degree, upon the

identification of diverse elements that make up the argument in toto, including the al-Mu'allaqat, or

collected pre-Islamic epics, and the Qur' än as evidence of this. This identification is also

problematic however, as the records of these epic poems are drawn from collections set down some

centuries after the apparent emergence of such a poetic lingua franca. These are highly refined texts,

that demonstrate a degree of perfection in their grammar, syntax and vocabulary, that experts

suggest is almost unique, and certainly makes it highly unlikely that they will have stemmed from

such a conflation of 7th century dialects. Lecomte suggests that the accepted position of recognising

that there exists some older material in what is essentially fixed, almost "stereotypical" specimens

of an older poetic tradition, incorporates the idea of apocryphal elements (Lecomte, 2003, vii: 254a

) but this is not the position adopted by some, and the assertion is altogether too certain. The idea of

the existence of an early poetic koine is also challenged by the accounts of gatherings where oral

epics were recited by differing tribal poets in a kind of competition. During these recitations, it was

apparently common for disputes to arise from misunderstandings, or a lack of comprehension about

style and language that strongly suggests no such common koine existed at this time.

To summarise, the form of Arabic that one might describe in the terms Rabin suggests (2003,

i: 561a), and that Lapidus and Hourani assert was in existence certainly amongst professional

reciters (räwis) of the pre-Islamic epic poetry (djähiliyya), namely a poetic koine, is problematic. It

is difficult to use as a precedent for the development of other such amongst conquered peoples, as

this is so specific a set of cultural conditions that it would seem unlikely to be able to apply it in

other circumstances. To support the argument for considering Domari as having evolved in

conditions that would seem comprehensible to those around this phenomenon, the Muslim Arabs, I

suggest a more common experience is necessary to identify (although this assumes that it is possible

to perceive the mentalite of peoples within a vastly different cultural and socio-historical matrix from my own, a conception that must be at least regarded with caution). The reason for pursuing

this consideration is that such an assertion by Islamic historians, or historians of early Muslim

society and culture, regarding the existence of this `lingua poetica', may apparently provide a better

precedent than that which I shall proceed to suggest. Despite this difference, it is nevertheless

109


instructive to outline some context for my suggestion that the development of Gypsy languages and

peoples, is intrinsically linked with the development of Muslim society itself and more particularly its military organisation, as an alternative to the view of Islam as alien to them, as frequently

alluded to in some Romani Studies scholarship.

110


3.2 "He created Man, taught him the mode of expression... " (55: 3-4) Arabic as a military koine

THE EXPANSION AND RAPID CONQUEST by the Muslims into the regions around the Arabian

peninsula, was achieved by groups of differing tribes-people, refashioned by faith and messianic

zeal that led them to dominate parts of the imperial lands of Byzantium (Egypt and Syria) and all of

Persia. This irruption must be seen as one of the most dramatic periods in human history, with

extraordinarily far-reaching effects stretching from the Indian Ocean to the Atlantic. It began, as I

suggested above with a profound readjustment of elements of Arabic society and culture in the

context of "ferment" and rapid social and economic change; in a sense it was a profoundly secular

and political readjustment as much as the outcome of revelation (see the works of Patricia Crone,

Michael Cook and John Wansborough). Its locus was the urban society of Mecca in the Hidjäz, and

the expansion of trade in the southern Arabian region that connected it to the Yemen in the south,

Syria and Iraq (Irak) in the north, and westwards to the Abyssinian lands, via the Red Sea. The

expansion of trade and communications, shifts in control of resources and in particular authority,

within the tribal groupings of the Arabs, especially the Quraysh, are the some of motors of change

in the complex web of factors involved. The other must be the person of the Prophet Muhammad

(peace be upon him), his visionary search for a secular and religious understanding of the world and

the altering that world in the light of revelation and what was a process of `nation-building'.

The evolution of the Hagarenes, the early adherents of the prophetic vision granted to Muhammad

(as they are described by Crone and Cook, 1977: 8-9), into Muslims was one that saw a

reconstruction of elements of Arabic society, or more particularly a certain section of Meccan

society in the period of the Blessed Prophet, the early 6th century CE. The ferment that Lapidus

characterises at the core of Arabic society is the basis from which the immediate group around

Muhammad developed (the community of believers or umma, a word whose complexity developed

along with the phenomenon it described; see Lewis, 1996: 53). The revelations that the Blessed

Prophet brought to the pagan, clan-based matrilineal system of the Arabs challenged the existing

institutions and notions of society, kinship and authority. Social structures, mores, religious beliefs

and economic organisation formed a whole, inextricably bound together (Lapidus, 1995: 25). As the

teachings of the Blessed Prophet developed, his oppositional position to the dominant norms and

values of the surrounding community became more acute and this is the impetus behind the

exclusionary treatment by the Makkan elites towards Muhammad and his followers (Hourani, 1991:

17) though the first eight years of the Muslim era show a pragmatic co-existence with pre-Islamic

elements, such as the alliance with tribes that maintained adherences to Manat, al-Lat and al-Uzza,

111


the supposed three daughters of Allah (Ali, 2003: 25-6). The removal of the fledgling community of

believers to the oasis of Yathrib or Yasrib, as it is recorded in the Qur'än (33: 13), some 200 miles

north of Makka, is the event that underpins the rise of Islam, the Hijra of 622 CE/I AH (Lewis,

1996: 52). The development of this community of sahäba, or Companions of the Blessed Prophet

(see Muranyi, 1994, viii: 827b), is a complex one that Crone suggests has been fundamentally

misunderstood as a result of the recasting of early Islamic history in the century following the death

of the Blessed Prophet (1997: 7-15). This community was however, moulded by his vision and the

experience of his leadership in al-Madina, where he had been offered the opportunity to take up the

role of arbiter in what appears to have been a deeply divided urban society, subject to factions and

almost constant feuding, probably as a result of economic competition (Lapidus, 1988: 25). A move

towards monotheism may be indicated by the use of the word Allah in some Madinan poetry, as a

designation for a supreme God (Watt, 2003, v: 995a), prior to Muhammad's arrival in the oasis. With

the compacts between certain of the tribes-people and the Blessed Prophet, the move by the

Hagarenes began in 622 CE/1 AH, with some seventy-three men and women from Makka preceding

the relocation of Muhammad and Abü Bakr, in September of that year. The organisation of the

defence of this new community against the attacks upon it in the following years can be discerned

from the "Constitution [Compact] of Medina" that it describes the inhabitants of the oasis as a

single umma (Watt, 1956: 225-6). This text would appear to be a composite one with many

repetitions and accretions, yet it is interesting in many ways here as it details the membership and

rights of the community, including the Jews of Medina and those who were not converts to the

religious message brought by the Blessed Prophet (Watt, 2003, v: 996b; Barakat, 1979). The process

of social organisation is one that underlies the text and it is in this process of reconfiguration that I

suggest we can identify the approach of the Muslims to the societal organisation, to the

incorporation of conquered peoples (Crone and Cook, 1977: 83-106).

The linguistic reconfiguration that occured in the conquered territories, particularly in the cities and

established military centres resulted in the evolution of a series of urban dialects that produced

innovations and differentiations not only between groups of differing confessional orientation

(Christians, Jews and Muslims), but also between classes and gender, often in the same urban

context (Gibb, 1986: 574). The primary mechanism for this change was the blending of the older dialects of Arabic into a "... sort of koine, rather military in character, which constituted the

language of the conquered or newly-founded towns" (Gibb, 1986: 574). It was this koine, with its

innovative differentiation between groups, that provided the impetus for the development of Doman

112


and the evolution of the Dom identity and a `model', or perhaps ̀ accepted' process for the evolution

of the Ghaznävid koine that, I suggest, underlies modem Romanes.

113


3.3 al-Sind, al-Hind and Rädjpütana in eighth century; the problem of the origins of the Dom

THE LINGUISTIC EVIDENCE DRAWN FROM DOMART clearly suggests an earlier departure date from

India for this group or groups (Hancock, 2005: 5-7), than the ancestors of Romani speakers, a

proposition that has been argued since at least the early 20th century (see Macalister, 1914; Fraser,

1992: 34-40). Potts' 1846 treatise may be the first scholarly consideration of the subject, although it

was based upon observations by an American missionary, rather than Potts' own research in the

field (Rev. Eli C. Smith, in parts of Syria). Matras' consideration of the contemporary picture in one

particular case, the Dom of Jerusalem, gives a linguistic model for the arguments of difference and

similarity between Domart and Romani, and the separate developmental path of the former (1999: 1-

58). The current series of practical and accessible Domar web-lessons by Donald Kenrick (2000-),

based largely upon Macalister's work, includes a piece entitled "What is Domart? ", where he states

"[it]... preserves the neuter gender for nouns whereas Romani has only feminine and masculine"

(2001). This last is the crucial juncture that is, in some aspects contested by some scholarship

(Matras, 1999: 7) and it is important to briefly examine this notion as it underpins the entire debate

and can be summarised as a question of linguistic origins and their relationship with the historical

group (this is, in a sense shared by the whole corpus of Romani Studies historiography).

The argument for considering the language spoken by those Gypsies in the Arab lands as an earlier

emigre from the sub-continent revolves around the transition between what linguists describe as

Middle Indo-Aryan and New Indo-Aryan, a process that Fraser suggests took place in the Indic

languages "during the centuries leading up to AD 1000" (1995: 18). That this shift in forms was not

complete, or that dialects maintained older forms for some time is agreed by most commentators

(Matras, 2002a: 30-34); what is sometimes at issue is the argument about the speed of such a

transformation (at a recent conference organised by the Roma Education Fund in October 2007,

Dieter Halwachs argued that the transformation is still going on). That Domari retains a form that

reflects the older construction of nouns, as Kenrick suggests (Kenrick, 2001), is again something

shared by many scholars although the current status of this form is not clear until further research is

published and existing research confirmed (see John & Hancock, 2008; Hancock, 2000: 1-13). The

arguments may be summarised as linguistic analysis having demonstrated that the dialects that

came to make up Domari must have departed the environment of Middle Indo-Aryan during the

shift in forms and the loss and reassignment of nouns in the neuter gender that took place c. 1000 CE

(Hancock, 2006: 77-78). This may be the same as suggesting that the ancestors of those groups that

have spoken and continue in some communities to speak Domart, left the orbit of the Indian lands

114


earlier than this point, though the link between languages and those speaking them may not always

be so clearly defined. The case of a number of languages of trade and exchange are a case in point

of fact; not all those who spoke Aramaic along the silk routes in the 3`d century BCE were of

Semitic origins, nor living in the Syrian and Mesopotamian lands. Arabic is another example, and

perhaps even clearer, as the language of religion, scholarship and trade that spread from Trans-

Oxiana to the Iberian Peninsula in the 9th century (Fuck, 1986: 569-71) and similarly Latin could be

described as such. The predominance of Latin in those areas of the Roman Empire where other

languages were the "mother tongue" ensured its continuing use as a lingua franca in the post-

imperial period. Most especially (in the context of this discussion), the essential place of Latin as

the language of the Roman military-administrative system, made certain that mercenary and

auxiliary troops of non-Roman origins learnt and used a form of the language (especially amongst

the officers of these units), as this would facilitate promotion, in addition to efficient

communication (see Wells, c. 1999). Koines had of course been part of the development of other

identities and to this extent, the connection between groups of Indians in the heartland of the

Khalif'a and the later Dom is not necessarily a unique one, but it is significant in that this took

place with a social reconfiguration that again had an earlier precedent in Islam.

The case for accepting the proposition that the two are indeed coterminous here must rest upon the

notion that Domart is a pariah tongue, without the kudos of carrying a new revelation, or being the

mechanism for literary exchange, or government as in the cases above. As the speech of a

marginalised and socially excluded and discriminated group, it is hard to see why the transmission

of it would have been anything other than through the group that used it and indeed, it is likely that

as a boundary marker of group identity (Barth, 1969), it would have remained deliberately limited

in its use. The important question of at what point the speakers of what came to be Domari left the

Indian lands must be considered as part of the question of how the language came into being. The

suggestions that speakers of the various dialects that formed the basis for Domani can be identified

categorically in one historical setting or another, is at the heart of the debate and part of the

problem, as the movements of Indian peoples has been a constant factor over millennia. From

Indian troops at Thermopylae (Kartlunen, 1989: 13), to the presence of HindU doctors in the time of

the Blessed Prophet, the incidence of Indian populations in the Persian lands, Byzantium, Arabia

and elsewhere in the region is attested. If we are to determine the origins of the Dom, it is necessary

to identify the most likely of circumstances in which they emerged.

A convenient starting point is to consider the suggestion of Hancock (2000: 9-11), regarding the

115


emergence of Domar-i as a koine outside of India, in similar circumstances to those pertaining to the

emergence of Romani in the llth century CE. Hancock begins by stressing the discrete and distinct

histories of Romani and Domara, "separated by over five hundred years" (2000: 9). Other authors

also have argued for the separate origins of the two (Matras, 1995: 29-59) and Turner's description

of what he defined as European, Armenian and Asiatic Romani first challenged the Sampson theory

of a single origin that then divided in the wake of a sojourn in Persia (see Fraser, 1992: 28-39, for

discussions about the doubts of identifying Romani and Domara as having common origins).

Hancock goes on to briefly outline the elements of shared lexical content between the two and

demonstrates that there is a very low incidence of Persian borrowings. "It [Doman]... is clearly of

an old-Indic type, and differs significantly from Romani in its phonology, grammar and lexicon. "

(2000: 9) This argument is further developed in Hancock's recent article (Marsh and Strand,

2006: 77-78) where a comprehensive list of words further supports his evidence. Having established

the underlying position of a separate historical trajectory for the two languages he goes on to

examine their different origins by starting from the account in Firdawsi, of the Indian musicians

arriving in 5t'' century CE Sassänid Persia. As I have argued elsewhere in this book, the historical

"fact" behind this legend is extremely unlikely though Hancock's linguistic analysis confusingly

concludes that the episode is "quite likely" to have taken place, but that "it does not hold up

linguistically" (2000: 9). This position is altered by Hancock's later emphatic assertion that despite

this argument being repeated frequently in recent scholarship "it is wrong" (2002: 5), as the

relationship between Romani and Domart is clearly different due to the evidence of a third gender

in Domara (2002: 7). Doman has its origins in a series of dialects that lexically are more similar to

others in the central regions of the Indian sub-continent, rather than the north western region where

the legend suggests these musicians would have hailed from. The origins of Dom populations,

Hancock goes on to suggest may be found in the groups of troops from a variety of backgrounds,

that left India to deal with the military threat of the Hephthalite Huns who invaded north-western

and northern Indian lands and eastern Persia, during the late 5 ̀h and 6`h centuries CE (2000: 10).

Citing a variety of historians (Thapar, 1966; Wolpert, 1977; Wright, 1969 and Watson, 1988)

regarding the "White" Huns, or "Ephthalites" (2000: 11), a link is suggested between the Gupta

Kings' defence of their territory against these and the origins of the Dom in a composite military

force that attempted to repel them. A further series of connections is posited between the early

mediaeval Hindü state of Kabul or Kapisha and groups of Dom in Persia in the mid-19th century

(Hancock, 2000: 11), although again as he suggests, it is not clear what this relationship is. It is

possible to see a direct relationship between Hancock's argument that the linguistic evidence for

116


dismissing the Firdaws1 legend as pertaining to the origins of the Romani peoples is the likelihood

is that these musicians would have been speakers of north-western Prakritic, whereas Domart shows

lexical similarities with central region Prakrit dialects (Hancock, 2000: 9). Thus, the characters in

the Firdawsi legend are likely to have been from Sind and not Hind, a suggestion that would

support the possibility raised earlier of Sindhi troops as allies of the Sassanids in their defence

against the Hephthalite Huns.

There are a number of issues with this series of suggestions that require some deeper investigation

before I go on to suggest an alternative narrative. The linguistic arguments regarding the similarities

with central Indian dialects and Doman are in the context of this overall book, beyond my own

competences and my argument refers frequently to the work of Hancock and other linguists to

support the historical picture presented throughout (whilst acknowledging that these themselves are

part of a developing narrative). Nevertheless, the relationship is not clear it seems to me, in

Hancock's own depiction of the origins of the Dom, as it relies heavily upon Thapar's suggestion of

the development of linguistic evolution during the Hunnic invasions. He cites her as remarking that

the shift from Prakrit (literally "natural" or "unrefined") to Apabhramshä ("falling down"; see

Fraser, 1992: 16-20 for a discussion of the shifts from Old Indo Aryan or Sanskritic, through Middle

Indo-Aryan or Prakritic to this most advanced form, before a final shift to New Indo-Aryan around

1000 CE), was partly the result of the influx of refugees and migrants from north-western

territories, fleeing into central and western India in the wake of the Hephthalite assaults (Thapar,

1966: 257, cited in Hancock 2000: 10). Initially, this would appear to draw an inference between the

dialect groups of north-western, central, southern and other Prakrits and a clear geographical location that is coterminous with these (Thapar, 1966: 257, cited in Hancock, 2005: 10). This is of

course not the case and the notion of a clearly definable territorial boundary between one group of dialects and others is altogether too neat. In this case, it is not clear if we are being told that the

north-western dialect groups (Sindhi, Landha/Western Punjabi; see Fraser, 1992: 18), impacted

upon the central and western Prakritic dialects (Panjabi, Rajasthani, Gujarati, West Hindi), or that dialects spoken by people from the north-western parts of India influenced those spoken by

inhabitants of central and western India, which would of course mean the dialects spoken by the inhabitants of north-western Hind impacting upon central and western Hind and not the speakers of dialects in Sind (north-western Prakritic). What this does suggest is the problems that arise from

linguistic's use of historical data and their assumptions about it.

In this case, it might appear that the displacement of peoples influencing dialects one upon another

117


would confuse the assertion that Domart can be traced sufficiently, in its origins to one group of

dialects to the exclusion of others, unless it is possible to relate the origins of Domart especially to a

particular linguistic context. If Domart is similar to Romani in its historical development in the

circumstances in which it arose, then the situation that pertains to the latter, namely

... that not enough is yet known of the emergence of the modem Indian vernaculars to take

the quest beyond rather general comparisons of shared evolutionary features which do not

permit [us]... to identify a particular language with certainty as Romani's closest relation...

(Fraser, 1992: 20)

would surely suggest a similar problem with Domart. The main point of distinction as Fraser

suggests, is the presence in Domart of a third neuter gender, traceable to the shift from Middle Indo-

Aryan to New Indo-Aryan, c. 1000 CE and an origin during the 5th century CE, but most discussions

suggest that Domart is more closely related to central Prakritic, than north-western, suggesting that

it is locatable to a particular context. These shifts are cumulative, and give an imprecise measure of

linguistic change, as mapping the linguistic development of any dialects relies upon being able to

corroborate, through textual analysis, the proposed path of development (I am indebted to Professor

Matras for what much of what follows, arising from a conversation 26th June 2005; any

inaccuracies remain due to my imperfect understanding of his points). The most ancient of

developments provide the baseline from which others follow, as a point in real time, and the

divergence of common vocabulary at a relatively constant rate allows for the "dating" of these

shifts, to a greater or lesser degree, by cross-referencing with examples found in contemporary texts

(see also Fraser, 29-32). Clearly, the relationship between the written and spoken word can be

profoundly different, and the status of recorded language as representative of elite discourse must

be taken into account. The essentially conservative nature of written texts, and examples of

deliberately archaic forms remaining in use in literature well after their disappearance in the spoken

idiom, must ameliorate any claims to pinpoint accuracy in terms of "dating", and a wide margin of

error has to be allowed. Similarly, the regional specificity of dialects is also elastic, and the notion

of a continuum, rather than hard and fast geographical boundaries between differing dialects, is far

more useful here. There are, however, "frontiers" where differences can be ascertained, and in this

way it is possible to see the origins of Romani and Domani as pertaining to the central region of Prakritic dialects (Panjabi, Rajasthani, Gujurati, Western Hindi), and their genesis in the period c. 9th centuries CE. Whilst a dynamic picture must be borne in mind when considering these points, however, it is may not possible to locate the notion of a group of "composite military troops moving

118


westwards out of India" to confront "an invasion of India in this area at this time" with the

ancestors of the Dom, nor to suggest that they may be related to "Indian troops [that]... left India to

engage them [the Huns]. " (Hancock, 2000: 9-10) The support for such a critique of this position

must be considered in the context of the broader historical picture, and it is to this that I now turn,

with an alternative hypobook.

The original migrations that led to the emergence of people we now call Dom were, I would suggest

forced and not voluntary, and took place as a result of the Arab invasions of Sindh in the early 8th

century CE and not during the earlier Hephthalite invasions. The notion that Hindus left India to

confront the Hephthalite may appear to find some confirmation in the texts of Firdawsi and al-

Hamza Isfah ni, in their locating an origin of the people they knew as Lüri in these earlier events

(see below), but I would argue that the necessary conditions for the fundamental reconfiguring of

Hindu groups are not to be found in 5th century Sassanid Persia, but in the complex and turbulent

Muslim societies of the 8th and 9' century Khalif'a. In order to develop this argument, it will be

necessary to briefly outline the course of Muslim history in these centuries.

119


3.4 The Umayyad and Àbbasid Caliphates: The Hindus into Dom

THE CAMPAIGNS OF MUHAMMAD B. AL-KASIM AL-THAKAFI in 711 CE/93 AH to conquer Sind, took

place under the patent of the Iräki governor, al-Hadjdjädj, during the Khalif a of al-Walid I b. Àbd

al-Malik. The assault of the Arab forces upon the region saw the Indus valley up to Multän,

conquered in fierce and bloody warfare over the three years of al-Thakafi's generalship. Enormous

numbers of Hindus were purportedly killed and captured (including 6,000 kafir prisoners at Räwar

and at Multän), but the key to understanding the subsequent evolution of the Dom identity, is the

incorporation of mawäli or client soldiers, into the Muslim Umayyad armies, as giqiniyya or

d4j4ý)L"l units (see Crone, 2003: 284-300; 1991: 874-882). These client troops remained Hindu

and brought auxiliaries (cooks, fletchers, tent-makers, sword-bearers, grooms and their own

attendants) and wives and offspring, captured at the battlefield in the attendant camps that made up

the majority of any mediaeval army. All armies at this time and well into the early modern period

were effectively societies on the move, as Hale has remarked (1989: 62-3), including merchants,

traders, scribes, bureaucrats, moneylenders, provisioners and others. Approximately two-thirds of

any army was made up of non-combatants, and most soldiers took wives and children with them on

campaign. Defeat then would mean the transfer of surviving troops, and their families to the victor's

side. The probable picture of the development of the Dom would seem to indicate that this military

group were decimated at points over a period of time, leaving groups of Indian descent without a

military function in Islamic societies, but maintaining the artisan skills associated with the auxiliary

and attendant personnel, especially following the Àbbäsid Revolution that ultimately deposed the

Umayyads in 750 CE. In the context of Muslim history, rebellions against the Umayyads have been

related to the discussion of Romani origins by Donald Kenrick (2004: 24-5). Kenrick argues that the

Arabic sources mention the Zott/Zutt an Arabic version of Jhät, as part of an earlier rebellion in

alliance with Sayäbiyya (Sumatrans), and suggests that these Zott were the progenitors of the Rom

(following de Goeje, 1903), after their defeat and capture by the Byzantines in their conquest of

Àyn Zarba, on the Syrian borderlands in 855 CE. White refers to the Zutt being removed to the

region of Antioch (Antakya in modern Turkey), under the Umayyad Khalif Mu'äiyya ibn Abi

Sufyan (661-680), to colonise the area of Suwaidiyyah (modem Samandag) in the Orontes delta

(White Jr. 1974: 201 - 221). These Zutt and their Sayäbiyya allies had joined the Àlids (followers of

Ali ibn Abi Talib during the First Fitna, or `time of trials') in the early civil wars of the late seventh

century (656 - 709 CE), andwere therefore suspect in their allegiances to the Umayyad Khalif'a as

Shi'ia Muslims. The continuing disturbances involving the Zutt in 710 CE would indicate that their

120


presence and adherence to the party of Ali remained at other locations, such as Mamistra (modern

Yakipmar, near Adana), another city retaken by the Byzantines in 964 CE and ultimately lost after

many changes in occupation, to the Ottomans in the 14th century (it remains a Roman Catholic

titular see: Vaihle, 1913). The attempted re-Christianisation by the Emperor Nikephoros Phocas in

the years following 964 CE are likely to have eradicated or removed any vestiges of the Zutt

population exiled there in the 8th century. These rebellious Zutt were also present during a much

larger revolt that threatened to topple the Àbbäsids in the later 9th century, led by one Ali

Muhammad (see below). Kenrick's suggestion that these groups transported onto Byzantium (at

least as regards the Àyn Zarba Zutt) form the basis for the Romani in Europe is problematic, as the

linguistic arguments outlined above would demonstrate that this is too early; it is possible that such

a group may have contributed to the presence of proto-Dom communities in the Byzantine Empire,

though this is speculative. It would not provide a possible suggestion as to the origins of the Dom

peoples of the Arab lands however; for this I would suggest the later turmoil that erupted in the

Àbbasid Khalif'a provides a better explanation. In order to present this hypobook, it is necessary to

briefly examine the history of the Àbbasid period.

Though initially Shi'ia, the Àbbasid movement quickly shifted to a more orthodox Sunnism with

the reign of the Khalif al-Mansur (754 - 775 CE) and the emphasis upon the dynasty's relationship

to the Blessed Prophet through his uncle Abbas. Al-Mansur's son, al-Mahdi (775 - 785 CE)

introduced much of the Persianised bureaucratic practice built upon Sassanid models that became a

hallmark of the later Khalif'a, as well as extending the use of client soldiers and recruiting large

numbers of Turkish slaves or mamlüks to the Muslim armies from the 830's under al-Mütasim (833

- 842 CE). This in itself was to prove the undermining of Àbbasid rule, as these mamlüks

eventually grew powerful enough to create their own power-bases and render the Khalif' a impotent.

By the close of the reign of Harlin al-Raschid (786 - 809 CE) the seeds of conflict were in evidence

in the Àbbasid polity and civil war erupted upon Harün's death between his two sons, al-Amin and

al-Maimun, dividing the state between Iraq and Persia, but the Àbbäsids had already faced severe

revolts in Persia and had all but lost control of the Maghreb during the reign of Harlin. With a

renascent Byzantium on the offensive under the Emperors Nikephorus (802 - 811 CE) and

Theophilos (829 - 842 CE) in Syria and Asia Minor, and the challenge of the Khazars from the

Black Sea steppe in the Armenian lands, the Àbbäsids also faced the break-away of their Persian

military commanders such as the Thurids in Khorasan (820 CE), to be swiftly followed by others. The use of the Turkish mamlüks did little to resolve the problem of diminishing Àbbasid authority,

121


as these too followed suit and established nominally loyal principalities including the Ghaznävids in

the 970's. The seat of the Khalif'a became insecure and al-Mntasim transferred this to Samarra in

836 CE, further alienating and weakening Àbbäsid control, and did not return to Baghdad until 892

CE (Hourani, 1993: 22-43).

Religious conflict also played a significant part in the demise of Àbbasid power, as the Khalif al-

Maimun attempted to enforce the adherence to the Mütazilite school upon the religious

establishment, something of a return to the original Shi'ism of the dynasty, after 813 CE and until

the accession of al-Muta'aqil in 847 CE. This policy was extremely unpopular and many of the

rebellions faced by al-Maimun and his successor al-Mütasim were a response to this. The

reassertion of Àbbasid control that marked the reign of the Khalif al-Muta'agil attempted to curb

the power of the break-away Persian and Turkish military commanders, but his murder in 862 CE

effectively forestalled any success in the long-term and eight years of anarchy followed until al-

Muä'faq became the effective power in the Khalif'a, while his brother al-Mütamid held the

Khalif'a (Lapidus, 2002: 56-66). However, even during this period of attempted imposition of

central authority, the fourteen-year rebellion of the Zanj or Zang severely challenged the hegemony

of the Àbbäsids, and brought it close to collapse.

The Zanj rebellion (869 - 883 CE/255 - 265 AH) broke out amongst the East African slaves of

Somalian, Eritrean and Ethiopian origins who worked building the vast irrigation projects that the

Àbbäsids engaged in at this time, in the marshlands of the Basra region as related by the historian

Muhammad ibn Jarir ibn Yazid ibn Kathir, Abu Ja'far al-Tabari (d. 923 CE) (al-Tabari, 1988-2007).

The rebels were reportedly led by a descendant of the fourth Khalif Ali ibn Abi Talib (son-in-law of

the Blessed Prophet), and therefore claimed to be fighting for the re-establishment of Shi'i

hegemony in the Àbbäsid Khalif a, as they asserted that the later Caliphs had departed from the

path of Ali and the first `rightly guided' imams of the past (see Popovic, 1999: chap. 2, who

disagrees with the notion that Ali ibn Muhammad was a descendant; the translation of the 1976

French monograph is problematic; see Morony's review article 2000: 1842). Ali as leader of the

revolt, brought together a formidable force that also included renegade mawlä (client soldiers),

Bedouin, Zutt, Suwaidiyyah and mercenaries from the Àbbäsid armies, creating an effective state

that reportedly had six or seven fortified cities that al-Tabari describes as impregnable and

established a capital in the Shat al-Arab region (Mesopotamia) called al-Mukhtara. The state was a

military one and the cities were organised around the production of weapons and warships, for use

upon the Tigris and Euphrates. During the height of the rebellion, the armies from al-Mukhtara

122


reached to within seventy miles of Baghdad, still the major city in the Àbbasid Empire despite the

seat of the caliphs being at Samarra at this time. The disaffection of the population from the

Àbbäsids was apparent in that there were fears that Baghdad would open its gates to the rebels and

join with them, and only rapid action by a swift-moving Àbbasid force to bolster the defences

quelled this in 877 CE. On a number of occasions the rebels defeated the Àbbasid armies sent

against them, and reportedly large numbers of the troops changed sides as a result after the battles.

The eventual defeat of Ali ibn Muhammad seems to have been as a result of betrayal and capture

through the incitement of al-Muä'faq who orchestrated a wholesale policy of amnesty and bribes to

various sections of the rebel forces in the later years of the revolt, and with Ali's execution or death

in the siege of al-Mukhtara (the sources are confused in their accounts; see Popovic, 1999: Appendix

III), the leadership of the rebellion collapsed and al-Mukhtara was taken and sacked in 883 CE by

al-Mud'faq, a date slightly later than al-Tabari gives as the conclusion of the rebellion, though

insurgence linked to the Zanj would seem to have continued and the Àbbäsids only reasserted

widespread control with their return to Baghdad as the heart of the Khalif'a in 892 CE and the re-

organisation of the state (Popovic, 1999: chap. 5). The destruction and dispersal of the Zanj saw

large numbers of former rebels reduced to slavery, and pariah status in the Àbbasid lands, or fleeing to the regions beyond their immediate control in the Maghreb and northern Africa. It is in

these groups that I suggest we find the ancestors of the Dom.

123


4.1 Koines and cavalry charges: the Kshatriya theory and the problem of Rajpüts and Jhäts in

Romani Studies

The argument that the ancestors of the Romani people, are the descendants of the Indian Kshatriya

warrior caste from the late 10th and 1lt" centuries has been posited for some time. In defining the

origins of the Gypsies, reference to Jhäts as the original "tribe" from which Romani peoples are

descended began as early as 1849, when Sir Richard Francis Burton (1821-1890, himself of

probable Romani origins; see Lovell, 1998: 27) suggested the connection in his essay about the

Jataki-Belockhi dialect (1849: 81-125), and went on to note the role of Abu l'Kasim Mahmüd b.

Sebüktigin of Ghaznä (969-1030 CE, Haig, 1993: 133; or 971 CE/361 AH-1030 CE/421 AH,

Bosworth, 1991: 65), in the fragmentation of the Jhäts and other tribes of northern India. Burton

dated this dispersal in the 11' century CE, in the context of his short reference to Mahmüd (Burton,

1849: 83). An earlier date for the dispersal of the Jhäts was suggested by de Goeje (1903: 20-33), in

relation to the expansion into Sind (al-Sindh), the region around the lower course of the Indus river,

by the "Omayad" (Umayyad) caliphs in the period 93 AH/711 CE, up to the Ghaznävid incursions

of Amir Sebüktigin (c. 365 A11/975 CE). The question of the Jhäts was also tackled in the

monumental work of Sir Henry Meirs Elliott, (1867-77). In an ethnological note to the first volume

(1867-77a: C, 507-508), Elliott quotes General Cunningham's Archaeological Report for 1863-64,

equating "... the Jäts [sic] with the Xanthii of Strabo, and the Iatii of Pliny and Ptolemy... " and,

suggests their origins are in the region of the Oxus river, amongst the Abars (Elliott, 1867-77a: C,

507). The designation of this generic term by him suggests a connection with the Avars or Juan-

juan, as Chinese sources record them (Sinor, 1990: 293). The role of the Avars in Turkish and

Byzantine history has been discussed previously in this book, here the possibility of their

connection with Romani history is also raised. This southern Indo-Scythian tribal confederation became known as the Jhäts, in contradistinction to their northern relatives, the Meds (Elliott, 1867-

77a: C, 507). De Goeje (1903: 20-33), and most recently, Kenrick (2004: chap. 2), have attempted to

link the Jhäts, or al-Zutt as they are recorded in the early Arabic sources (see Bosworth, 2002: 574),

to the Gypsies from their transportation from Sind by the Säsänid emperor Bahräm Gür V (r. 420-

438) to their settlement by Abü Müsä al-Ash`ari (614 CE-42 AH/662 CE), companion of the Blessed Prophet and governor of Basra, and eventual Byzantine dispersal from `Ayn Zarba after 220

AH/835 CE (Bosworth, 2002: 574). The Jhäts latterly appear in the annals of the Ghaznävid

assaults upon the region of Multan in 418 AH/1027 CE (Briggs, 1829: 47-48), when Mahmüd's

seventeenth expedition took place. He defeated a force of Jhäts who had taken to the waters of the

124


Indus, by constructing a substantial flotilla (reportedly fourteen hundred boats, although

Elphinstone and Elliott both express doubts about the figure or circumstances of this particular

action; see Elliott, 1867-77b: D, 465), armed with iron rams (Briggs, 1829: 47). The defeat of the

Jhäts resulted in most drowning, whilst their families, having taken refuge on some islands "... fell

into the hands of Mahmood [sic]" (Briggs, 1829: 47). The Jhäts also appear in the T rikhu-s

Subuktigin of Abü-1 Fazl al-Baihaki, c. 451 AH11077 CE, when he relates the story of Tilak the

Hindu's defeat of the rebellious Governor of Hind, Ahmad Niältigin in 426 AH/1035 CE Elliott,

1867-77b: III, 109-112). This episode illustrates that they were part of the auxiliary forces that the

Ghaznävid commanders could mobilise, despite the conflicts between the Jhäts of Jüd, and Sultan

Mahmüd in 417 AH, recorded by Nizämi-d din Ahmad's Tabakat-i Akbari (Elliott, 1867-77b: D,

466). Elliott also suggests elsewhere that the Jhäts are the descendants of the Yueh-Chi via the

Kators (1867-77b: A, 403-4) in a complex geneaology that takes us back to the central Asian Avars

and Hephthalites. The presence of these particular Indians amongst the larger Hindu community at

Ghaznä is likely from this point in time, but the genetic core of Romani ancestry is unlikely to be

exclusively, or even predominantly from this single group.

As I have argued previously, the Arab assaults on Sind are at the root of the question concerning the

origins of the Dom, and de Goeje's work (1903) may provide some explanation in identifying the

8`h century CE Muslim invasions as the catalyst in the process of Dom ethnogensis. In the context

of Gypsy history, the evidence that the Umayyad general Muhammad b. al-Käsim al-Thakafi

captured many thousands of Indian prisoners (both warriors and non-combatants; see Haig, 1997:

632; Fredunbeg, 1900: Section 8) may offer an explanation of the ancestry for the Domari-speaking

Gypsies, though again the need for the chroniclers to glorify the victories of the Muslims over the

infidel Hindli should be borne in mind. The linguistic evidence would certainly not support such

and early departure date for the ancestral Romani population, as the evolution of Old Indo-Aryan,

and the loss of the third neuter gender, did not take place until some three centuries later (Hancock,

2004: 8). The kshatriya theory has had supporters since these first suggestions (Leland, 1882;

Kochanowski, 1968), most recently Hancock (2006; 2004; 1987), whose arguments combine both

linguistic and historical analysis to advance the case. However, not all have been, or indeed are

convinced and others continue to propound the notion that the origins of the Gypsies lies in related

castes of itinerant traders, metalworkers, entertainers and snake-charmers associated with military

expeditions or caravans (see Hübschmannovä, 2004). Indeed, the scepticism of Fraser (1992: 45),

when discussing the Kshatriya theory has been taken further by others, who deny any validity at all

125


to the notion:

"... in a number of recent publications, Hancock claims that Romani was formed as a

military koine by a caste of warriors assembled to resist the Islamic invasions of India. In

some circles, this view is gaining popularity as it pretends to revise what is referred to as

potentially racist, or at least stereotypical images of the Rom. There is, however, neither

linguistic nor historical evidence to support it" (Matras, 2004: 301)

Matras has elsewhere argued in the past for a view similar to that of Hübschmannovä, namely that

Romani is spoken by "descendants of itinerant castes of artisans and entertainers who are spread

throughout Central Asia, the Near East and Europe. " (2002: 1) The most recent text by a scholar of

the "social constructionist" school also remains extremely sceptical about the idea of the Rajpüt

origins, in the context of an overall criticism of an Indian ancestry, suggesting that any linguistic or

historical "evidence" is at best imprecise and unclear (Mayall, 2004: 223-224). As others from a

similar theoretical perspective have suggested, the kshatriya or warrior caste theory of origins

reflects something of a Romani political aspiration (see Willems & Lucassen, 2000), that can be

seen in contradistinction to the earlier Gypsy-lorist identification of origins amongst the lowest and

most despised "pariah" groups in Indian societies of the past. It is in sharp contradistinction to this

earlier view of origins, that some Romani Studies scholars declare their agreement with Hancock's

perspective (Lee, 2004: pc; Marushiakova & Popov, 2001; Kenrick, 2004).

The arguments surrounding Romani origins are, as suggested earlier, always a complex mixture of

scholarly debate, linguistic analysis and defence of particular positions that may reflect wider

political and social agendas and this particular book is no exception. The question of why in this

instance it has remained the subject of constant enquiry and sometimes heated debate is in itself, an

aspect of "the problem of Gypsy identity" and proceeds from the "contextual, constructed and

contested" nature of such (Strand & Marsh, 2006), as Mayall has so eloquently demonstrated (2004:

219-251). However, much of the establishment of contra positions to the kshatriya theory suggests

that Hancock's own argument lacks any earlier reference, that it is in fact sui generis. An

examination of the development, or reiteration of this notion in his most recent text (2005), clearly demonstrates that it has preoccupied those concerned with the question of exactly which group of Indians Gypsies actually descend from (to somewhat oversimplify the proposition), for some one hundred and fifty years. Hancock's arguments are clearly part of this historiography, which he

explicitly details (Hancock, 2005: 3-6).

126


Persistent myths of origin are, of course, the staple of many ethnic identities, and indeed the basic

building blocks of many national communities. In this sense, those attempting the construction of

Romani identity are following a well-worn path:

... if Elizabeth of England was the heiress of pre-Saxon Britons, if the Swedish kings were

descendants of the Goths, the French kings sprung from Gaulish loins and the Muscovite

tsars (through a particularly weird conceit) related through Ruirik to the Emperor Augustus,

then it was not too eccentric for the Polish commonwealth to boast of its origins in a race of

Iranian `barbarians' from the Black Sea. (Ascherson, 1996: 231; for a discussion of Polish

`Sarmatism' and nations as ̀ forged' communities, see Ascherson, 1996: 230-6)

One might add that it seems not too improbable that Gypsies might claim to originate from al I'h

century warrior caste of disparate elements, frantically (and heroically perhaps), defending their

towns and temples from the onslaught of an especially militant Islam. Yet of course, it is this very

notion, that of the heroic defeat, the military catastrophe in the face of overwhelming odds, that has

been addressed recently in a challenging proposition that these moments of "death and glory" are

part and parcel of the development of many political and "ethnic" communities (Schivelbusch,

2004: 25) and in particular the Hindutva historiography discussed earlier. In addition to providing

the foundation for societal and moral recovery and being a major motivation for the propagation of

myths of an inverted cultural superiority, the vanquished are forced to re-examine their identity, and

to reconstruct their own history to explain the disaster that has befallen them (Schivelbusch, 2004:

35). In some of the examples above, the experience of defeat and conquest by a people deemed so

utterly alien to those rendered ̀ subjects' by the conquerors, not only provided (and still continues to

provide) an explanation for questions of `backwardness' and under-development, it has also

underpinned major social and political change in the societies that underwent this catastrophe. The

rise of Muscovy, and the development of the Romanov imperium can be seen not merely in the

terms framed by nationalist historians determined to prove any perceived inadequacies of their

nation-state upon the ravages of the so-called "Mongol-yoke", but also in respect of the

reconfiguration of Russian identity and society at crucial points (the reigns of Ivan Groznii and

Peter the Great for example; Matuschevski, 1984; Anisimov, 1993). The question might be that if it

is not improbable that Gypsies could make such claims to a warrior inheritance, is there any support

for this proposition? Could one examine the development of Romani identity as outlined below,

with the notion of their response to unremitting defeat and cultural destruction of large numbers of

Indian people and their socio-cultural environment, in mind? If one can accept Schivelbusch's

127


argument that military disaster produces unprecedented psychological change in human societies,

that shifts in identity follow such catastrophes (a suggestion that McCarthy has also posited in

connection with the ethnic cleansing of Ottoman Muslims in the wake of Russian victory in 1878,

that we shall return to in examining Balkan Gypsies in the 19th century; 1995: 227), the evolution of

a distinct language and culture in the Ghaznävid milieu that resulted in the eventual coalescence of

Romani ethnicity seems distinctly possible. Despite the certitude of Matras (2004: 301), or the

scepticism of Fraser and Mayall (1992: 45; 2004: 223-224), there are, I would suggest clear

historical probabilities for such an assertion.

In considering such a question, it is essential that one assess, from a critical position vis-a-vis

"history", or "History", any "evidence" and "facts", before attempting to propose a plausible basis

for this position. As a Romani historian, it is my intention to outline in this section the historical

material that I argue supports the contention that Gypsies can in part can be said to stem from

ancestors who fought as warriors with differing military statuses, both in the armies of the Rajpüt

princes who attempted to repulse the advance of Islam under Mahmüd b. Sebüktigin (387-421

AH/997-1030 CE), and his son, Mas'ud b. Mahmüd (421-431 AH/1030-1040 CE), and in the

Ghaznävid forces of the same period. I further argue that such elements that did not originate

amongst the warrior caste, derived from the remaining auxiliary and associated groups that

inevitably, and invariably accompanied mediaeval armies in this period and were likely to have

been in the majority, given the impact of the subsequent defeat of the Ghaznävid forces in 431

AH/1040 CE at Dandängän. The havoc wreaked by the destruction of the elite warrior groups

would have rendered their numbers much lower, weighting these in favour of those that were not directly involved in the three days of bloody struggle and the eventual calamity of defeat on the

battelfield. This analysis, that problematises the overly polarised positions adopted by individual

protagonists to date establishes the historian's more common position with regard to all such

questions, namely that the debate thus far fails to recognise the complexity of the factors involved.

These have been hidden by the simple-minded dichotomous presentation of warriors or pariahs as the historical `parents' of the Romani peoples. I have attempted to present a more elaborate picture

of Romani origins in the difficult and sometimes contradictory interplay of 11 ̀h century northern

and northwestern Indian Raj put principalities and the Ghaznävid Empire that dominated what we would now describe as Afghanistan, Trans-Oxania and Khorasän. In short, it is in all these

groupings that one must seek to find any kind of resolution to this discussion, not merely as a means of assembling some compromise that accommodates both positions to achieve some sort of

128


consensus, but precisely because of its historical plausibility. Patently, an explanation for the origins

of Romani identity and ethnicity that suggests a single factor, or military group as the primary

genetic core (as it has been described that Hancock suggests, though this is not the case; see Matras,

2002: 201; Hancock, 2006: 84), cannot be considered tenable, any more than one that denies such a

possibility as part of the picture overall. We must seek for a explanation of composite social and

occupational origins (Hancock, 2006: 70), in addition to the genetic variation identified by

Kalaydjieva, Gresham and Calafell (2001).

129


4.2 Ghuldms, mawlä and àbid; the enslavement of Hindus in the Ghaznävid state

WE KNOW THAT THE GHAZNAVID SOVEREIGNS ORGANISED INDIANS among their troops, as we know

from the Tärzkh-i Yamini that Mahmüd was in the practice of doing with all conquered "nations", as

exemplified in his treatment of the Khiljis, Afghans, and others (Elliott, 1867-77: II, 28-29). The

ancestors of the Romani peoples I suggest, were formed in the complex socio-psychological

environment of being conquered Indian àbid (slaves, singular àbd, albeit with a particular

economic and military importance), under Ghaznävid control and could only have come together as

a result of external pressures that compelled otherwise distinct castes to respond to the unique

situation they found themselves in. This was to identify the commonalities that separated them from

the surrounding Muslim communities in the experience of slavery and servitude to the Ghaznävid

dynasty (in a similar way to the Turkish mamlük soldiers of the Àbbasid Khalif a), rather than

maintain the internal divisions that marked them off from each other. This is not to suggest that such

differences in status were rendered null and void. Clearly those Hindus that were converted to Islam

prior to their capture (and there must have been some from the Ismaili Sind principalities that were

the target for Sebüktigin and Mahmüd's initial raids in the early 400's AH/1000's CE), together with

any that adopted the new faith as a consequence of their enslavement and the possibility for their

manumission, must have been to some extent set apart from the majority of their compatriots,

however suspiciously regarded or condescended to by their Sunni masters. The status that

Ghaznävid society, itself highly militarised, ascribed to the elite warrior group and the palace

ghuläms, may also have had an impact upon this corps in their dealings with other Indians,

elevating them above the common-or-garden infantry, men-at-arms and, almost certainly the

auxiliary personnel that supported them (see the illustration below for a reconstruction of the image

of Hindu ghulams).

The status of the troops that fought from elephants, one of the most prized elements of the Ghaznävid army (they are consistently mentioned in the narrative sources amongst the booty, and

on one occasion, awarded an honorific when at Kanauj an elephant belonging to the Raja Chandar

Räi made its own way to Mahmüd's camp and was called Khudädäd, "the gift of God"; Elliott,

1867-77b: 453), may well have had some impact upon the social standing of this group. Entertainers, dancers and musicians probably did hold some social value for both their community

and their rulers. It is recorded that these were present on the occasions when the Ghaznävid sultans and their immediate companions sought respite from the normal run of activities in drinking parties (majälis-i nashät u sharäb), at which Turkish epic poems were recited, musicians, dancers and

130


clowns performed (Bosworth, 1963: 138). Given our knowledge of the pattern of households in

other Islamic polities (Hathaway, 1996), the likelihood that such entertainers were sought after by

those emulating their leaders in the upper echelons of Ghaznävid society is high. In some ways it

may have been these individuals or groups of entertainers that found it most easy to cross the

barriers of status, religion and ethnicity, catering as they probably did to both the Ghaznzvid elites

and the lower classes, in the Persianised courts and the Hindü mahalles. Merchants and traders who

are recorded as accompanying the Sultan Mas'üd on his disastrous Dandängän campaign in 1040

CE, must also have been present in the Hindü community in Ghaznä and on Mahmüd's earlier

campaigns and razzias 997-1030 CE (Bosworth, 1973: 139-40 for a descriptions of the Ghazndvid

bazaars). Their status was probably similar to that of other conquered mercantile populations; as

'abid they may have been entrusted with a high degree of autonomy, frequently running their

master's businesses and often having dependent clients or mawlä of their own (see Crone, 2003:

290). There were also the `invited' sons and daughters of HindU rulers who owed allegiance to the

Ghaznävids at the sultan's court; these occupied a privileged, though sometimes precarious position

as part of the elite in Ghaznävid society. Finally, it is unlikely that the common wives, children,

camp followers and domestic slaves were regarded especially highly by the wider group, given the

emphasis placed upon martial qualities and the degree of military organisation in the society at large

and the importation of Persian-Arabic cultural norms as regards the seclusion of women (Bosworth,

1973: 138). In this instance, the class distinctions that existed previously to the capture of large

numbers of Hindus and their relocation to the centre of the empire, were possibly further reinforced

by the attitudes of their conquerors to this group, as defeated and non-military 'abid. The position of

the auxiliaries and military servitors was probably much higher; Mahmüd held his Hindu engineers

in high regards as essential contingents in his campaigns, often relying upon them to reduce cities

and opponents rather than resorting to a pitched battle (Bosworth, 1961: 112). Skilled artisans, of

whom there were literally thousands involved in the building of palaces and the "Bride of the

Heaven" mosque in Ghaznä, may also have had a socially elevated position above the mass of

domestic and military labour.

In order to provide the basis for a more informed, and ultimately more tenable `narrative' outlining

the particular context of the ancestry of Romani people, in relation to both the notion of the

kshatriya warrior caste and the more nuanced picture of a reconfigured, militarised society in 11th

century Ghaznä, it is necessary to discuss the record that does exist for Hindü Indians in the empire, in some detail. The problems of examining the Ghaznävid Empire revolve around the issue of

131


sources and accessibility. Those such as al-'Utbi's al-kitäb al-Yamini (AbU Nasr Muhammad b. 'Abd

al-Djabbär, d. 427 AH/1036 CE or 431 AH/1040 CE; see Bosworth, 2000: 945; Elliott, 1867-77b:

II), the Tärikhu-s Subuktigin of Baihak (Elliott, 1867-77: III), or the Tarikh-i Firiýta (see Briggs,

1829; Dow, 1768-72), provide much narrative material (and this should not be underestimated as to

its worth in modern historiography), but what else exists that might corroborate the suggestions

outlined here offer far less detail than might sufficiently support these propositions, leaving what

many could argue are the most unwarranted of speculations. It is in order to contextualise these that

a comparative picture has to be drawn, one that allows a degree of `reconstruction', or the

suggestion of the most likely of outlines for the book that I have proposed here. The main engine of

change in constitution of originally divers Indian populations was the defeat and forced migration

of differing groups of Hindü people to Ghaznd, where they were reconstituted as a new militarised,

composite group and forced to develop both new linguistic and cultural formations in response to

the enormous changes they faced. The imperial heritage for this group, as for so many others in the

mediaeval period and since, provided the basis for the emergence of a new ethnicity in the 12th

century, an identity that has remained with us to the present as the Rom.

The institution of slavery in all Islamic polities was, of course not completely without nuance in the

relationship of master and slave (see Lewis, 1994: chap. 1; Brunschvig, 1986: 24-40), but given the

numbers of those enslaved and the decline in the price of individuals in this state during this period,

it is hard to imagine that an especially beneficent attitude prevailed on the part of the conquerors,

even when they were moved in individual cases to manumission upon conversion to Islam. In the

context of those defeated by the Muslim forces, their status as captives and domestic or artisan

slaves was altered by manumission to that of mawlä (freed men and retainers; see Crone, 1991: 874-

882). This process was a common feature from the Sufyänid Khalif a onwards (660-683 CE; see

Crone, 2003: 287), originating in the tribal protection offered to ethnic and occupational "pariah"

groups in pre-Islamic Arab society (Crone, 1991: 874). Though most mawla were converts, being

non-Muslim was not necessarily an impediment to the validity of a mawlä relationship (Crone,

1991: 876), or to holding positions of responsibility (Brunschvig, 1986: 3 1). Once a mawäli, [sing. ]

the opportunity arose for payment as military combatants (Crone, 2003: 287), or to follow other

occupations and trades. The appearance of regiments of these client soldiers in the Muslim armies

was increasingly a feature of the post-conquest Islamic forces, especially after the 2"d Civil War that

undermined the Arab monopoly of military power and authority (Crone, 1991: 877). As discussed in

the previous section, the Hindü Qigäniyya regiments were conscripted as soldiers in their own right,

132


along with Berber Waddähiyya and Iranian Bukhäriyya who provided special corps with particular

skills (Crone, 2003: 287; 1991: 878). Significantly, contingents were often led by mawäli who

themselves claimed non-Arab and aristocratic origins (Crone, 1991: 877). I would suggest that we

can see this heritage lasting indirectly, also influenced by the impact of the sojourn in the Armenian

lands with the arrival of groups of Gypsies in western Europe in the early 15`h century, that clearly

reflect this type of organisation (Fraser, 1992: 66).

In the context of the organisation of Muslim armies and in particular the forces of Ghaznä, their

composition and structure retained this status of some Hindes as mawlä regiments hierarchically

structured, as distinct from those who maintained their `abd status. Whilst their status and pay was

never equal to that of the mugätila or `regulars', they were enrolled, together with their sons, slaves

and clients of their own to carry forward ghazä or Holy War, sometimes with little choice in the

matter (Crone, 2003: 287). The complex question of the movement of Hindü groups from north

western and northern India into the region of Khorasan and Persia proper can be exemplified in this

process; recruited as ̀ native' regiments of converts, or conscripted after capture, enslavement and

manumission, their deployment was into the frontiers of the Muslim territories to carry forward the

conquest of new territories for Islam, in those situations where the use of groups originating in these

lands would be avoided because of local or tribal loyalties. The deployment of Indians in the

western empire's territories was a deliberate policy to ensure no local ties obstructed the will of the

sultan (Bosworth, 1963: 107) and that "the logic of the system seems to have been that people were

recruited from those furthest removed from the establishment" (Lucassen & Zürcher, 1999: 5). Thus,

the consistent perception of Hindus and Romani people as resistant to the Muslims must be

ameliorated by the evidence of their participation in the advance of Islam for some several hundred

years, as soldiers and supporters of the faith in the Ghaznävid and Ottoman states. I suggest that this

perception has a great deal to do with the need for 20th century Hindi influenced, Rroma

nationalism to identify anti-Muslim groups who resisted and were dispossessed by the Islamic

invasion and incursions (Thapar, 2000). The subsequent coterminous development of Romani

evangelism and political organisations, with the leaders of the former extending their influence and

authority into the latter, identifies an ̀ intrinsic' Christian identity as part of a modern definition of Romanipe (the Romani `way') and seeks to perpetuate this notion, sometimes explicitly (Strand,

2001: 47). The opposite notion, when presented to groups of modem Pentecostal Roma evokes a

strong response ("I understand this, but I don't like it" as a Finnish Kalo once remarked to me).

Conceptually, the notion that one can define with certainty the absence of any historical "evidence"

133


for the proposition of a warrior caste origin for the Gypsies relies upon a number of positivist

assumptions about the nature of such evidence, and then proceeds to argue from those assumptions

as if they indeed represented the totality. It also ignores the possibility that groups in the Islamic

world were both soldiers and servitors, even slaves. As I have outlined earlier, this wider

epistemological position is one that has become not only untenable amongst historians in the past

twenty or so years, it is inconsistent with the current discussions regarding history as an expression

of collective identities, and suggests that a more "scientific" approach in fact exists (that which

underpins linguistics, in this instance) and would clearly identify through the analysis of a particular

lexicon, the difference between groups of people with differing statuses and functions in any past

society, in an uncomplicated and accurate manner. Such an approach would it seems (to present one

example), allow us to differentiate between the various "components" that make up the Mongol

population of the present to determine those who descend from the warrior elites, those who stem

from those groups that provided what might be termed "auxiliary services", and those whose status

clearly separated them out as slaves, concubines or camp followers.

The comparison is not accidental; the organisation of the Mongol state, a highly militarised form of

reconfigured tribal structures, represented the maximum effective military power available to

cingiz Khan and his descendants that made it possible to mobilise almost the entire society outside

of the conquered, sedentary populations for warfare. The highly stratified nature of the Mongol

military machinery (based upon the tümen or decimal system of multiples of 10), recognised an

elite surrounding the person of the khan and an imperial guard of 10,000 cavalry made up of individuals from a wide variety of tribal backgrounds, bound by ties of personal loyalty. This

structure included commanders of decreasing status depending on the force they led, the tamma

contingents (stationed as border units on the perimeters of the steppe-sedentary limes Mongolica)

and the locally recruited infantry and garrison troops (see Morgan, 1993: 232-233; 1986: 78). The

Ghaznävid state in some ways prefigured this organisation under Sebüktigin and Mahmüd. The

degree to which one can support this particular comparison will be detailed below; the point here is

that it would not be possible to categorise the present Mongol population into groups upon the basis

of being descended from either the elite warrior corps (the imperial guard), the more general cavalry or the large body of tent-makers, fletchers, armourers, cooks, herders, and others making up the

support structure of the Mongol armies of the past. I would argue on the basis of such a comparison (and there are many others that could be made), that the absence of particular language structures that allow a clear definition of the descendants of a warrior elite within the modem-day, identifiable

134


speech-community (as linguists such as Labov suggest exist; Gordon, 2006) of Mongols, cannot be

taken as the basis for stating the historical absence of such. To concentrate the point still further, any

shared military terminology that exists in modem-day Mongol-Turkic clearly cannot derive only

from groups that may have been merely "associated" with those for whom these were concepts with

a concrete reality, rather than an awareness of the objects they represent.

Any apparent linguistic concentration around, or presence of military terminology does not and

cannot demonstrate in and of itself proof positive of the social, military or political status of

individuals or groups of speakers in the past. What this does do (as with all historical "evidence"),

is to offer an opportunity to further support what might be the historical context for the fragments of

the past that remains to us. Clearly Hancock's position as it has been most frequently presented

(though not by him), posits a plausible relationship between the lexical "borrowings" from a variety

of Indic languages, deriving from Middle Indo-Aryan forms, that are identifiably "military" and

present in modern Romani, with an original Rajpüt warrior caste in the 11'h century. If this were

indeed the heart of the debate it would prove short-lived, as obviously any language that contains

military terminology in its modern form cannot be said to demonstrate that all the speakers of that

language currently descend from military personnel. Undeniably however, some do and others may

well count from their ancestry those that were amongst the elite formations of any military

organisation in the past, however distant from those individuals in the present. Elements of the

present-day, Mongol-speaking population are clearly the descendants of the warrior elite from the

past whilst others are not (the persistence of ethnic and `tribal' groupings in modern Mongolia

maintains these divisions; see Keyser-Tracqui et al, 2006: 272-281). To suggest otherwise would

surely posit an unbroken heritage of military duty in families or groups going back centuries, which

of course exists in some cases, as the only legitimate corroborative of the linguistic presence of

terms for weaponry, manoeuvre or equipment in the modern lexicon. The argument that an

awareness of the concepts for "sword" or "spear" amongst a current "speech community" (arguably

as imaginary as any other community), fails to prove social status in a past military (or militarised)

society, cannot disprove that members of the present community that share these terms in the wider

context of a definable language or dialect, are not indeed descended from those that did once hold

or acquire such status. As I have argued earlier, the development of any war idiom takes such

terminology well beyond the confines of military personnel and into the apparently non-combatant

sphere. The use of a wide range of military terms in common parlance in specific situations during

wartime, is an aspect of the dramatic changes caused to language in such conditions.

135


4.3 "Speaking to the scribes, talking to the troops... " The development of the Ghaznävid koine

IN THE CONTEXT OF THE PRESENT DISCUSSION, there is significant continuity in the approaches of the

chanceries of the Ghaznävids, the Saldjükids and the Ottomans, to the question of a need for a

language in particular to be used amongst the military and scribal classes that come to significantly

differ from the "common tongue" of the wider society. It is especially interesting that all three

polities resolved the problem through the incorporation, either wholesale or in part, of Persian as the

language of government and military command. Whilst Arabic remained the language of

jurisprudence and religious writing (and to some extent of continuing exegesis, with the

development of sufism), the impact of Iran upon Islamic civilisation fundamentally altered the

direction of its growth (Crone & Cook, 1997: 56), and nowhere more so than in the realm of

governance. The Ottomans developed a version of Turkish using an Arabic script and infused with

much Persian and Arabic that effectively created an unbridgeable gulf between written and spoken

forms (Kramers, 1993: 918). The active promotion of Ottoman culture and arts, as a means of

strengthening their particular identity vis-ä-vis the Byzantines and the development of Ottoman

Turkish as a language in its literary and scribal forms, were characteristic of both their admiration of

the renaissance of Persian culture under the Timürids at Herat and Samarkand (Tekin, 1995: 211),

and the highly competitive atmosphere of the various Anatolian principalities in the earliest period

(1300-1453 CE). As a world empire at the opening of the 16th century, these literary concerns were

suitably imperial, political and worldly. The translation of older Islamic and pre-Islamic texts were

frequently undertaken, often through adaptations or translations of religious works, medicinal texts

or collections of love stories popular all over the Islamic lands (Nizämi's Khusraw wa Shirin for

example c. 581 AH/1184 CE; see Mu'ayyad, 1977), and narrative epics such as Firdawsi's Shäh-

näma, (c. 401 AH/1010 CE) was a prime concern of a number of Anatolian begs and emirs as a

means of establishing a particular court culture and wider literature. It also helped to standardise the

kind of language used, as did the production of dictionaries that translated terms between Cagatay

and Ottoman, for example. It is beyond the scope of my book to examine this particular point more

closely, but it is important to establish the fluidity and malleability of the differing Turkic languages

or dialects during this mediaeval period (Tekin, 1995: 210). The development of what has been

mistakenly labelled as "Old Ottoman", and more correctly known as "Old Anatolian Turkish", is the

product of the evolution of Turkish after its entry into Asia Minor, under the political patronage of the Mim Saldjükids, but the language of the court remained predominantly Persian although the descriptions of the Saldjük Sultan Muhamad b. Mahmüd's (511-525 AH/1118-1131 CE) drinking

136


parties record the fondness for his hearing recitations of Turkish epic poetry (Luther, 2001: 126).

The Saldjükids adopted Persian as the language of chancellery and court, almost as soon as they

had displaced the Ghaznävids' control of Khorasan and Iran, in the wake of the collapse at

Dandängän (431 AH/1040 CE; Bosworth, 1963: 250). Although initially using the surviving

Ghaznävid officials and scribes to communicate their victory after the three-day battle to the wider

Islamic world, the great vizier Nizäm al-Mulk (408-485 AH/1018-1092 CE) undertook to create an

effective administrative machinery that he outlines in his Siyasat-näma, or "mirror for princes" (al-

Mulk, 1956; Darke, 1960). The Ghaznävid machinery of war and administration, the latter being to

an important degree an arm of the former, were incorporated into the Saldjük polity without

significant alteration to the basic structure, thanks to the continuity of people like Nizäm al-Mulk,

whose family had been part of the administrative class that survived the vicissitudes of Ghaznävid

state-building (Bowen, 1995: 69-73). Taking the previous Ghaznävid state organisation as an ideal

model, and in particular praising the multi-ethnic composition of the army (Kai Kä'üs b. Iskander,

writing in the late 11th century notes, "he [Mahmüd]... constantly overawed the Hindus by means of

the Turks and the Turks by means of the Hindus, with the result that both nations submitted to him

through fear of each other", Levy, 1951: 153; Bosworth, 1963: 107), Nizam al-Mulk's advice for his

Saldjükid masters makes it clear that the languages of government, religion and literature were

Persian and Arabic, not the rough Turkish they had brought with them from central Asia as the tribal

Oghuz or Ghüzz. He also notes with approval that "the army should be composed of troops of all

races" and organised by nationality separately under their own commanders (al-Mulk, 1956: 107),

again supporting the notion of a koine developing that would facilitate command and

communication in such circumstances. The prestige of Persian in the eastern Iranian world had

brought about a state of affairs that can be best described as bilingualism, with the continuing use of Arabic in Ghaznävid culture and the chancellery influencing those that defeated them (de Blois,

1995: 970-972). Indeed, the continuity of governance of Khorasan, Herat and other regions of the

Perso-Islamic world was a key factor in the changes in rulership during this period; protecting the interests of merchants and peasants was of utmost importance in securing support in the Iranian

heartlands (Bosworth, 1995: 955). The Ghaznävid failure to do so in the face of the Turkmen

ravages convinced the population of this region to desert the dynasty for the Saldjükids, as better

able to control this unruly element (Bosworth, 1995: 955). The impact on the Saldjükids of these

complex negotiations was to intensify the process of sophistication that was they were undergoing

and further increase the influence of Persian as the preeminent language in the discourse of power.

137


In these three cases, the linguistic configuration of the elite underwent significant changes

throughout the administration and military spheres; in the Ottoman case the shift was from "Old

Anatolian Turkish" to Ottoman over two centuries, with profound influences from Persian and

Arabic as the Ottomans confronted the Byzantines and Greco-Roman culture. In the case of the

Saldjükids, the language of state and government increasingly became Persian as the move from

nomadic tribal confederation to Muslim power-state took place, whilst Arabic dominated the

religious discourse as the dynasty moved in to the place vacated by the Ghaznävids, as defenders of

the faith. In Ghaznä, the shift was, I argue, reflected in the initial orientation of the state westwards

towards Khorasän and the early use of Persian (as with the Saldjükids), and latterly eastwards

towards India, but nevertheless the development of the military koine that underlies the emergence

of Romani identity suggests not only the use of Persian as the language of the military and

government, but a clear separation from Arabic as the language of religion and religious literature,

that would only be accessible to those that had converted to Islam.

The fact that Mahmüd maintained the Indian contingents as non-Muslims is attested in to the bitter

complaints of the anonymous chronicler of Sistän (see below and Bosworth, 1963: 89,110). Clearly,

the conversion of these Indians to Islam was not countenanced by Mahmüd and his encouragement

of the Karämiyya sect in Khoräsän, to persecute the Ismäilis, suggests a rather more complex

personality than that represented by the title awarded him at the opening of his reign in the official

laqabs (Haig, 1993: 133; Bosworth, 1973: 185-9). What is most important here however, is that the

exposure to the language of Islam and of Allah's revelation for his chosen people, was not

permissible to those who were not of the ummä. This was a result of the highly stratified Ghaznävid

polity and the absence of any Arabic loanwords in Romani does not necessarily reflect the more

frequent assumption that the ancestors of the Gypsies passed through territories as yet untouched by

Arab Muslim culture or influence. Despite the acceptance by scholars that Romani itself has been

and remains to some speakers, a language to be restricted to a particular group (or even age range,

as acquisition can be a rites de passage amongst some groups), the notion that Arabic was limited

outside of the Arab lands to those who were a people chosen by God and part of a community of

believers, seems to be unconsidered. The absence of Arabic in the Romani lexicon is an illustration

of the degree to which the Hindus experienced a separation from the surrounding religious

community, the ummä, indicating that as soldiers and auxiliaries in the Ghaznävid forces the

Indians maintained their primary religious orientation without converting to Islam (except in a few

probable cases such as the Hindu commander Malik bin Jai Sen, described in Ahmad's Tabakät-i

138


Akban; Elliot, 1867-77b, II1: 4). This was both as a matter of deliberate policy on the part of the

sultans, in ensuring loyalty to their persons and avoiding the establishment of any local allegiances

that might compromise this loyalty and as an effective means of the control of rebellious local

Muslim and Christian populations by regiments that could be expected to act punitively, as a result

of their enforced participation in the ghäzwarä (Holy War) against their quondam homelands (the

suppression of the revolting Christians and Muslims in Zaranj for example; see Bosworth, 1963: 89,

110). This last point is redolent of the earlier argument for seeing the development of the Romani

ethnicity as stemming from psychological dislocation and the necessary reconfiguration of identity,

resulting from catastrophic and consistent military defeat (Schivelbusch, 2003; McCarthy, 1995).

The savagery of the suppression of the Zaranj and Sistän rebellions can be seen as a consequence of

the frustration of the Hindus sent to deal with it; in an act of displaced revenge against those who

first defeated them, and then forced them to fight for the continuing "Holy War" of their Ghaznävid

overlords, the Indians unleashed their fury upon the Muslims, Jews and Christians of the city.

The development of the Ghaznävid koine represents both a means of efficient command across

ethnic and class boundaries, in the context of a multi-ethnic army and a vehicle for the Hindez

groups from a variety of differing petty-states and principalities to communicate across dialectical

differences. The incorporation of a significant Persian lexicon must be understood in the context of

Ghaznävid administrative and political structures, where Persian was the language of the state in the

wider Islamic world. In the context of 11`h century Ghaznä, the control of language was one more

mechanism for detaching the Hindii contingents from their wider spiritual and social community,

ensuring a dependence upon the person of the sultan for patronage and protection. As his loyal

servants, the Hindus could expect to be rewarded directly and maintained in spite of criticism

against the use of pagans and idolaters in the armies of Islam (Bosworth, 1963: 110). The later

Byzantine army, with its Varangian Guard originating from differing Scandinavian and Anglo-

Saxon ethnic communities, was in some ways analogous with this picture of the Hindu regiments of

the Ghaznävid army. Separated by language and culture, the Varangians maintained an elitism

(partly through being expensive to join), a distinct code of discipline, dress and battle equipment,

were physically distinctive and often notoriously loyalty to the person of emperor (Kazdhan,

1991: 2152) and were part of the military re-organisation of the imperial forces into the tagma

structure under the direct command of the emperor, as opposed to the regiments under the control of

the strategoi from the themes. The Byzantine Emperors developed these regiments that were

detached from the surrounding Greek population, court and administration, to act as a counter-

139


balance to the interests of the bureaucrats, and the military comes and dux, in the struggles for

power and control that frequently beset the empire (Norwich, 1993: 87). The most important

difference was, of course, the status of the respective formations; the Hindus were captured slaves

and remained so in the Ghaznävid structure, whilst the Varangians exercised an independence of

action (choosing not to fight upon the occasions of palace coups, for example), as paid soldiery.

Another example, perhaps closer to the Hindus would be the Tourkopouli of the Byzantne army,

captured Turkish children raised in the tenets of Christianity and trained to fight under their own

officers (Vryonis Jr. 1956: 433n. )

The koIne that became Romani, has been identified as being closest to Urdu in its structure and

lexicon (see Hancock, 2005: 7-20), and it is in the context of the Ghaznävid military formations and

structures that we must look for the relationship between them. The basis for their development is I

would argue the same, in the military koine of 11th century Ghaznd. The process by which this one

lingua franca separated is to be found in the events of 1040 CE, when the Saldjükids not only

defeated the Ghaznävids, but detached the western empire entirely. This resulted in the fleeing of

those troops who remained after the defeat (the ancestors of the Romani peoples) and their complete

loss of contact with those that had been on duty in the eastern territories. As both these sections of

the army would, I suggest here have shared the kolne, the trajectory of its development was such

that two distinct branches emerged over time resulting in Urdu in the eastern lands, and Romani in

the territories of Byzantium and the west. It is not possible in this book to examine the development

of Romani more exactly and it may be that a task for linguistic analysis of the language in the future

would be to test more thoroughly this proposition. Historically the suggestion that I have outlined

here may represent a plausible alternative trajectory of development for both languages in their

origins, in the context of the Ghaznävid milieu and given the subsequent separation of differing

speech communities. However, without further support from Romani linguistics' scholars such as

Hancock and Matras, this remains a speculative picture at best.

140


4.4 State and structure; palace and chancellery in the ghazi emirate

IN SUMMARY THEN, THE POSITION OF HINDU ̀ abid and mawlä in Ghaznävid society was similar to

that of any mamlük groups in Muslim polities. Some would have had the opportunity to rise in

status and found their way into the very heart of the Ghaznävid military and administrative

structures (as the case of Tilak the Hindü general during the reign of Mas'ud b. Mahmüd

illustrates). Those and the palace ghuläms would have held rank and honours in the militarised

society that the Ghaznävid state represented and maintained their position by dint of the patronage

of the sultän, and were dependant upon him for their survival. In return, they were deployed not

only on the numerous campaigns against the Indian rajas and princes that resisted Ghaznävid

invasions, but also against the rebellious subjects of the sultan who balked at hardening centralised

rule and increasing tax burdens. In this role, the suggestion I have made is that the complex tensions

of remaining a defeated, dislocated Hindii in a militarised Muslim society, being used in campaigns

against their original homelands and against their former places of worship and holding a status

that, for some was possibly elevated as warriors and may have been concomitant with their former

place in Hindü society (though clearly only a minority of those captured by the Ghaznävids) though

reaping the benefits of these campaigns, meant that nevertheless they were a conquered people,

enslaved and despised as infidels or heretics (if they had been Ismailis previously). The barbarism

that was visited upon the rebellious subjects of Zaranj and Sistän, all `people of the book' as the

anonymous chronicler complains, may have been a venting of these tensions, an opportunity for

retribution as the Zanj rebels had earlier sought retribution from the conquering Muslims in the 9th

century ̀ Abbasid state.

Furthermore, the Ghaznävid state, like the Ottoman state of the 14`h and 15th centuries was a

military state, a state whose entire resources were channelled into prosecuting war and securing the

plunder that resulted in order to sustain itself. Not all the Hindus captured by the Ghazndvids were

engaged in direct combat but practically all who were not sold as personal slaves were at the service

of the state and the vast military machine that Sebüktigin and his heirs constructed. Military

terminology was part of the daily life of these people as an idiom that permeated every aspect in

ways that went beyond the mere necessity of command and order. For a militarised state such as the

Ghaznävid Empire, the war idiom would have been the normal means of communication, from the

court to the streets of the cities themselves, war must have imbued the very breath of the

inhabitants.

The importance of the division of the Islamic political formation, in both its forms as diwän-i a`lä

141


(chancery) and dargäh (palace) underpins all structures of eastern Muslim states in a continuous

tradition until the collapse of the Ottoman Empire in the 20th century (Barthold, 1928: 227). The

intensely personal notion of power in many of these polities meant that there was often significant

overlap between the chancery and palace bureaucracies, so that this picture is something of an over-

simplification as the division of competencies of officials shifted according to various state and

individual rulers' needs, and in the case of the Ghaznävids, the distinctions between the two were

frequently blurred. In achieving the necessary level of co-operation between the bureaucracy and

the military, competencies were rarely so clear-cut as to allow us to accept Bosworth's rather rigid

description of Ghaznävid government unreservedly (1963: 82). It is reasonable to agree with the

more general arguments made about state, society and military, especially in the context of

innovation as it is arguably as a result of military necessity that bureaucratic reform takes place in

any mediaeval or early modern polity (Aksan, 1999: 23-34). The state's need for those who will

prosecute war on its behalf whether by conviction or coercion, has been and remains one of the key

factors in the relationship between the polity and the population (Lucassen & Zürcher, 1999: 8). The

ability of mediaeval and early modern Islamic states to mobilise their resources for military

expansion or defence was formidable, not least because the bases for the creation of effective

bureaucracies necessary to support such military formations and societies were incorporated with

the conquest of Iran by the Arabs in the 7th century CE. The period after the disruption of the 2"a

Civil War saw the acceleration of the centralisation policies of Abd al-Malik (65-86 AH/685-705

CE) (Lapidus, 1988: 5 9-65; Wickens, 1976: 71-77) and the "classical" development of the Muslim

polity and government. From this period, the development of an efficient dtwän (chancery) was an

essential prerequisite to the successful expansion of any fledgling Muslim principality or emirate (an established political and military structure led by an emir or commander; Lampton, 1983: 332). I

would suggest here in a broader context of the political-military development of Muslim Persia,

Mesopotamia and Asia Minor, that those emirates and border beyliks (generally frontier polities

centred upon a bey, beg or tribal leader, such as in the Saldjükid and Ottoman cases), that failed to develop this element of state structure were ultimately absorbed by those that did (the examples of the Denizli emirate, c. 1261-1278 CE, the E*refogullan in the Begshehir area c. 1250-1325 CE, the Sahibogullari in the region between Kutahya and Ak*ehir and the Pervaneogullari at Sinope

illustrate this point as they were all relatively short-lived and eventually absorbed by the Ottomans;

see Vryonis, 1971: 138). Paying for the army drove rulers to develop bureaucracies that could organise and efficiently levy taxes, or secure sufficient surpluses from merchants and the peasantry to be able to do so. The Ghazndvid solution to the problem was in itself unusual; very few Muslim

142


rulers had access to the sort of wealth that could be provided by raiding India on a regular and

consistent basis and in a sense, the military capacity of the empire was both a product of and a

response to the opportunities offered by the Indian states' resources. In economic and military

terms, Mahmüd's vast standing army was necessary to extract maximum plunder from the Indian

lands, but could not be sustained without those resources and the ability to dragoon large numbers

of troops into the army from subject populations across the empire. As Bosworth notes, a little

ironically, "... it was in his time an institution geared to the exploitation of India" (1963: 255). It was

a structure that Mas'üd b. Mahmüd (421-431 AH/1030-1040 CE) and his heirs could not maintain;

the loss of the western empire meant the reservoir of potential recruits for the army from outside

India was gone. Given the relocation of the empire's focus from Ghaznä to Lahore that the sultan

attempted after 431 All/ 1040 CE, the need to conscript personnel had to be met from the territories

that were much more proximate to the new heart of the empire, with the consequent disruption of

loyalties and outbreaks of rebellion from individual emirs and rajas able to mobilise their own

forces locally. Structures of command that were maintained in Ghaznd could not longer be imposed

by dint of geographical separation and cultural distance. Mahmüd's success was in disrupting the

patterns of loyalty and allegiance of all his troops and assimilating them to the structures of the

Ghaznävid state by creating a bias towards India without becoming embroiled there. Over time this

pattern was reversed as the relocation of the empire forced successive sultans to compromise this

model, whilst they themselves became more enmeshed in the regional relationships of power and

influence. In effect, the empire became an Indian, albeit Islamic polity rather than the Turco-Persian

state it began as. Although the notion of the ghazi state as one exclusively mobilised for the

prosecution of war has been taken to extremes in the analysis of Islamic empires and emirates (see

Kafadar for a discussion about the importance of the ghazT book in relation to the rise of the

Ottoman state, 1995: 38-59), the impact of this principle nevertheless had a profound effect upon Ghaznävid, Saldjükid and Ottoman administrations throughout their histories.

The razzias themselves were the main source of the early Hindü troops, booty and slaves for the

empire. In describing the numbers of captives, the narrative sources available to us suggest the scale

of the enslavement of defeated peoples. The conquest of Peshawar (392 AH/1001 CE) and Tähnesar

(402 AH/1011 CE) resulted in apparently enormous numbers of captive Indians (see the discussion

above). The descriptions from the aftermath of the conquest of Kanauj in 409 AH/1017 CE (where

Mahmüd allegedly took 53,000 captives, according to the Tabakät-i Akbar ; Elliott, 1867-77e: XL,

179), is that in addition to his own Indian troops, Mahmüd could rely upon those of his "allies", in

143


this case the Räi of Kanauj Kooowur Ray (Briggs, 1829,1: 40), sacrificed to the vengeance of the

Hindu confederacy. Killed for his allegiance to the sultan by the combined forces of the Chandel

Rajpüt Rajas, Pür Jaipäl and Nandä, the sultan undoubtedly used the former Räi's troops in his

defeat of both at the battles of the Jumma, and of Rähib during 421 AH/1021 CE (Elliott, 1867-77b:

D, 456). The later examples of the reign of Ma'süd b. Mahmüd (see below), show that the reliance

of the Ghaznävids upon their Hindu contingents continued unabated.

144


4.5 The ideology, military organisation of the Ghaznävid state and the Hindus

THE HISTORIC "MISSION" OF THE GHAZNAVID POLITY was the prosecution of ghäzwarä, as a

legitimating ideology of the state and dynasty (Bosworth, 1977: 8). In his message to the Qärägänid

Ilek Khan in 390 AH/1000 CE (Briggs, 1829: 29; although Bosworth suggests some confusion over

the date, see 2001: 166 n. 9), Mahmüd emphasises his role as ghazi "[w]e continually undertake to

wage holy war in the regions of Sind [sic. ] and the remote areas of India" (Luther, 2001: 30). It is

not my intention to oversimplify the significant differences between this ideology and the reality of

Ghaznävid patronage of the ascetic Karrämiyya sect in Nishäpür (as a means of overawing the

Àlids, the Ashàri-Shäfiì ulamä and the Sufis; see Hurvitz, 2003: 31; Bosworth, 1990: 669;

1973: 163-205), their own Hindu commanders and officers, or even Mahmüd's complex motivations

in patronising that most famous poet of at his court, the Shi'ia Firdawsi. The banner of "Holy War"

was one that most Islamic states raised in their justification for conquest and military expansion,

even at the expense of other Islamic rulers (Bayezid Yildirim's conquest of the Central Anatolian

beyliks in the 1390's CE, for example; Vryonis, 1971: 138). The highly militarised and centralised

Ghanznävid state was clearly of a different order than the much more diffuse, tribal and patrimonial

early Saldjükid polity centred upon Khoräsän, which ultimately maintained many of those traditions

despite its later sophistication and transplantation (Bosworth, 1986: 939). Yet both used the

legitimating claim of defeating the infidel (or heretic) in their promotion of the dynasty and its right

to rule. Both also governed peoples for whom the notion of descent from the family of the Prophet

(peace be upon him), or an affiliation with the subjects themselves had provided a key to the

effective assumption of power and authority, yet neither could claim such. The Ottomans in many

ways moved from the ideology of ghäzwarä as the sole legitimation for the activities of the dynasty,

to a more complex doctrine of imperial rule reflecting the inheritance gained from their conquest of

the Byzantine Empire. The awarding of titles reflecting the image of warrior for the faith was important to Mahmüd and his successors; the notion of the ghazi was at the heart of the Ghaznävid

sultäns' self images, and their conception of the state. Legitimation lay in the subjugation of the infidel and, perhaps more importantly, the heretic.

The key to sustaining the Ghaznävid enterprise however was to meet the demand for fiscal

resources to support the state and army, and those resources derived almost entirely from campaigns into the Indian subcontinent. They were essential to the florescence of the state and its continuing

expansion, and the ideology of ghäzwarä legitimated and justified the campaigns of Amir

Sebüktigin (366-387 AW942-997 CE), Amir (later Sultän) Mahmüd (397-421 AH/997-1030 CE),

145


and their successors in plundering the wealth of India. These were promoted to the wider Islamic

community, which in return provided further legitimation through the awarding of titles like "the

right hand of the Khalif 'or "defender of the faithful" (Yamin ad-Dawla wa-Amin al-Milla). Al-

`Utbi describes the receipt of this honour thus

Kädir bi-llah Amiru-l-müminin, the Khalifa of Baghdad, sent a Khila't [sic] (a variety of fine

garments, embroidered with inscriptions or tiräz; Stillman, 1986: 6-7), such as had never

before been heard of, for the use of Sultan Saifu-d daula, and he entitled Mahmüd in his

imperial rescript [decree], "Yaminu-d daula Aminu-1 millat, the friend of the Amir-1

müminin, " which had not yet been bestowed upon any prince, either far or near,

notwithstanding their intense desire to receive such an honour (Elliott, 1867-77: II, 20)

There is an illustration (above) of Mahmüd donning the khila't surrounded by the ulema from an

MS of Rashid al-Din's Djämi al-Tawärikh in Edinburgh University Library (Stillman, 1986: 6-7).

The relationship between the dynasty and the ulema or ulamä [Ar. ] had originally been one of

legitimation and support; the appeal to Sunni orthodoxy had been bolstered through the patronage

of the ulamä and the support of the Khalifa by defeating the Shiite Btiwayids and reclaiming

Baghdad (Lewis, 1995: 94). This, together with the establishment of medresse (madrasas Ar. the

earliest in Islamic history), under the Ghaznävids, allowed the dynasty to claim a legitimacy

detached from ideas of consanguinity with the family of the Blessed, ethnic, or regional claims

(Lapidus, 1966: 345). The Karrämiyya were frequently accused of heresy in their pursuit of a more

ascetic Islam, yet they can be better understood as part of what Lewis calls "the anti-Shiite Sunni

revival" (1995: 95). Yet the complexity of Mahmüd's personality, alluded to earlier, is to be seen in

the apparent contradictions to this imagery and ideology. The poet Abu `1-Kasim Firdawsi was initially encouraged by Mahmiid in the production of his epic, and the much bruted account of his

inadequate reward may owe less to the meanness of the sultan, or Firdawsi's implicit Shr'Tsm, as

much as to the association between the poet and Mahmüd's first vizier, Abu ' l' Abbas who was disgraced shortly before the poet arrived in Ghaznä (400 AH/1010 CE). The wider considerations of Mahmüd's beliefs and complexities as a character must be left aside, but the question has to be

borne in mind when we consider the actual place of Indians in the Ghazndvid system, and the

challenges to the orthodoxy of the antagonistic timbre of Muslim-Hindu historiography, as

suggested earlier in this book.

The administration of this system needed to be efficient and effective in the collection of revenue. The inheritance of the Säminids was invaluable, as many of the previous scribes and paymasters

146


remained in place after the succession of the Ghaznävids to their territories after 366 AH/977 CE,

maintaining the mechanisms necessary for the distribution of the igtä istiklal (or iktd' istighläl Ar. )

dispensations to the army (Lapidus, 1966: 346). The nature of the land-holding regime under the

Ghaznävids is not entirely clear, nor is the participation of the bureaucracy in this system, it being

much better documented for the Saldjükid era, but it does appear that the means of supporting

troops by providing revenues directly at source, continued throughout the Ghaznävid period

(Lapidus, 1966: 346). At some point during the reign of Mahmüd this system no longer functioned

with its previous efficiency and abuse became commonplace, so much so that the alienation of the

population of Khoräsän becomes widespread, through excessive taxation and the rapacity of the

Ghaznävid revenue agents (Bosworth, 1963: 242). The Saldjtikiydn were able to capitalise upon this

in their conquest of Ghaznävid cities such as Nishapür, during the latter part of Ma'süd's reign;

Ghaznävid rule was doomed, not by the defeat at Dandängän in 1040, but by the the ulamä

and ghazis of Nishapur, Marw and other cities who received the Saldjuk leader Toghril Beg

as their new Sultän... (Lapidus, 1966: 346-7)

The grievances of the ulemä, merchants and traders of Nishapür were expressed when Tüghril Beg,

the Saldjuk Sultan, issued his proclamation protecting the citizens of the city in Ramadan (June-

July), 429 AH]1038 CE. "Tüghril Beg came to Nishäpür and sat upon the throne of the realm of

Sultan Mahmüd", ordering his troops to avoid having "anything to do with the inhabitants of the

city, and the peasants, and let no-one molest a living thing. " (Luther, 2001: 37) The consequences of

this were disturbing indeed to Mas`üd, who launched his campaign against Tüghril Beg and the

Saldjüks that ended at Dandängän, but the implication in the Saldjük-nama is clear; the Ghaznävid

treatment of their subjects had deprived them of their loyalty in Khorasan due to excessive and

abusive taxation (Lapidus, 1966: 346).

Unlike the Saldjük and Ottoman polities however, the ideology of conquest was not a major part of

the undertaking for the Ghazndvids in India. It was a source of slaves who were both domestic and

military, craftspeople and artisans, entertainers and labourers, many of whom were forcibly

removed from their homelands and original domiciles throughout the northern Indian regions and

resettled as the Sultan's servants in the empire's centres at Bälkh and Ghaznä. It was a source of

political kudos and further propoganda of the pre-eminent position of the dynasty as warriors for

Islam, especially against the heretics of Sindh, but most important it was a source of fabulous

wealth, of riches untold in elephants, spices, silks, precious metals and jewels. The notion that the Ghaznävid Empire was an "army with a state" is an extremely apposite one (Bosworth, 1963: 98);

147


more than any other Islamic polity since the inception of the period of Muslim expansion in the

seventh and early eighth centuries, the Ghaznävids represented the ultimate application of the

principle of ghazwärä to the organisation of a state. Society was mobilised to support the

prosecution of war, and the state controlled production of many vital goods, being the primary

consumer of many others. Distribution was largely aimed at servicing the needs of the state, the

court, the army and the bureaucracy (Longworth-Danes, 1993: 153). The infrastructure of

communication and services was built and maintained almost exclusively by slave labour to ensure

the rapid deployment of troops along campaign roads into India and Khoräsän, with the

maintenance of trade routes frequently secondary to the efficient movement of mounts and men

(Bosworth, 1963: 118).

The army under Mahmüd b. Sebüktigin was the most highly professional of its day. The ghuläms or

slave commanders and palace elite, were the core of the military structure, commanding units that

were frequently ethnic in composition and often specific in function or specialism. The Arabs and

Kurds for example, were grouped under one commander and usually formed the bulk of the

vanguard during marches, with a prominent role on the campaign in Marw al-Shähidjän (Bosworth,

1963: 111). The Indians were commanded by their own Sipahsälär-i Hindüyän and occupied a

distinct quarter of Ghaznä, or mahalle (Bosworth, 1963: 110). At the heart of the Ghaznävid forces

were the mamlük (slave) elements, comprised of Turks, Tajiks, Khoräsänians, Dailamis (a mountain

people from the Caucasus region), and more importantly for our purposes here, large numbers of

Indians. These were captured, or received as tribute from outwith the empire and often trained as

converts, but with the Indians this was usually not the case (see the Tabakät-i Akbari for an example

of captives and converts being taken by Mahmüd in Kashmir 404 AH; Elliott, 1867-77: D, 446).

The reasons for this will be discussed more fully, but the Indians together with the Turks provided

one half each of the sultän's household ghuläms that numbered between 6,000 (under Mahmüd

c. 421 AH/1030 CE) and 4,000 troops (under Ma'süd b. Mahmüd c. 431 AHJ1040 CE) (Bosworth,

1963: 105). Given the structures of mediaeval Muslim Hindü or Christian armies, this figure will have represented about one third of the actual number of troops involved, as each elite warrior will have had a number of lightly-armed retainers, groom, shield and sword-bearer, armourer, tent-

maker, cook, and body servant, bringing the total figure up to three times that of 4,000-6,000 palace

ghuläms. In the sultan's household alone during the period of Mahmüd, the numbers of Indians

could have been some 15,000 troops and retainers in total. Baihaki notes that at the battle of Dandängän (421 AH/1030 CE), there were some 16,000 Türkoman, each with four other family

148


members (Bosworth, 1986: 938). This ratio would appear not to have been unusual, if we consider

the figures from other, though admittedly later, sources (see Vryonis, 1971: 260-267), and we might

conclude that there were somewhere in the region of 64,000 Türkoman present upon the Saldjtik

side.

The provisioning of Ghaznävid armies required lengthy baggage trains, strings of elephants and

numerous camp followers, all in addition to the actual combat troops (Bosworth, 1963: 249). It is

therefore likely that the armies of the sultäns were larger than the 10,000 or 20,000 warriors

described in the accounts of campaigns of Waihind, Nagarkot (399 AH/1008-09 CE), or Mathura

and Kanauj (409 AHi1018 CE), and included the attendant auxiliaries, merchants, wives and

children, sweethearts, prostitutes (both those officially "licensed" and other, more opportunist sex-

workers), washer-women, laundresses, cooks, foragers, carriers and nurses for the sick and

wounded. As Hale notes, "... armies, as populous as all the major cities, were self-contained

societies on the move. " (1989: 62-3) Women were a major part of these mobile "nations", carrying

weapons and, in the case of Turkish nomadic groups, effectively mounted warriors in their own

right (Vryonis, 1971: 261). The example of the bahdjyan-i Rum, the Türkoman women of Anatolia

mentioned by the Ottoman historian A§ikpa5azade that carried out military duties along with their

maternal concerns (see Angiolello, 1873: 94). if we accept the ratio as suggested by Bosworth, the

numbers of people in Ghazndvid armies might have amounted to 80,000. This figure my also have

been close to double on the return from successful campaigns, with slaves and captives included. In

the instance of the defeat of Djaipal at Peshawar, 8t" Muharram 392 AH/27th November 1001 CE,

al-`Utbi records a figure of five hundred thousand slaves, "beautiful men and women" (Elliott,

1867-77: II, 21), although Firishta suggests that these were later released, upon payment of a large

ransom (Briggs, 1829: 30)

Ghaznävid infantry, unlike its predecessors, was a permanent force and the first non-mercenary,

standing army of any size and sophistication of the Muslim world. Firishta notes that Mahmüd

fielded 30,000 foot soldiers (along with 12,000 cavalry and 300 elephants), for the campaign

against the Raja of Lahore (Briggs, 1829: 29). The Arab armies of the conquest period had been

based upon the tribal levy and the mawäli retinues of Arab chieftains. The later developments of the

Khoräsäni army of professional mercenaries under the Abbäsids "... brought an element of technical

reform which had been missing in the Umayyad army. " (Cahen, 1983: 505). The predominance of

the Turks in Muslim armies, as mamlGk, rather than `abd, was a significant development in the

replacement of the Khoräsanis and the advancement of Islam in general (see Lewis, 2001: 86-102).

149


The dominance of the Turkish element of Muslim armies up to the advent of nation-states, was

never seriously challenged after this. The remedy for unruliness and disorder amongst the troops

was to counter-balance one ethnic element with another (Cahen, 1983: 506). The mountain peoples,

such as the Dailimis, Kurds, Africans and Hindus were all deployed to achieve this balance between

various elements. The Qäbüs Näma (A Mirror for Princes) of 475 AH/ 1082 CE, praised the use of

this method for controlling both elements in the Ghaznävid forces whereby Mahmüd "... constantly

overawed the Hindus by means of the Turks and the Turks by means of the Hindus, with the result

that both nations submitted to him through fear of each other... " (Iskandar, 1951: 176). The Siyäsat

Näma of Nizam al-Mulk (408-85 AH/1018-92 CE) also favoured the Ghaznävid model, in a chapter

entitled "The army should be composed of troops of all races", he describes how soldiers should be

organised by ethnicity and encamped under their own guards on campaign to encourage

competition in battle between the different "nations" (Bosworth, 1963: 107). The ethnic diversity of

the Ghaznävid army was more than just a way of utilising particular "native" skills, it was a

mechanism of sophisticated "divide-and-rule", much favoured by later imperial administrators such

as the British (Bosworth, 1963: 107), though he does not suggest the incorporation of Hindus into

the Saldjük forces.

The Indians resembled the Turks in that they were divorced from their origins and apparently

maintained no ties of loyalty to their "home" (though, as evidence about the Ottoman dev D irme

system has shown, connections with the original families and kinship groups of these palace

külwere frequently maintained; see Kunt, 1983). As captives they were especially separated from

the wider population by both their ethnicity and their belief system, as pagan Hindus. Their

presence during the suppression of rebellions was the cause of outrage amongst Muslim

commentators; the anonymous historian of Sistän (now in western Afghanistan) complains bitterly

of the violence and slaughter caused to Muslims and Christians in Zarang (Zaranj) by Mahmüd's

Hindu troops in 393/1003 (Elliott, 1867/2: 59-60,130-131; an illustration from an MS of Rashid al-

Din's Djämi al-Tawärikhin in Edinburgh University Library shows the campaign, and depicts an

Hindez siege-engineer in the lower left-hand corner of the miniature). Because of this absolute

loyalty to the person of the sultan, Bosworth suggests that the Hindus were regarded as more

reliable, and certainly during the turmoil of the period following Mahmüd's death in 421 AH/1030

CE, when the succession was moot, the Hindu troops commanded by one äy, remained loyal to

Sultan Muhammad b. Mahmiid during his brief reign (1963: 110). This proved no badge of dishonour for Sultan Mas`üd Abü Sa'id, who eventually came to control the empire. Mahmüd's

150


eldest son relied even more heavily upon the Hindus of the empire's forces in his efforts to roust the

"old guard" of his father who had shown him little loyalty initially, and used the Indians as a prime

means of enforcing his will against partisans of his brother, fifty days after his father's death,

according to al-Baihak (Elliott, 1867-77b: III, 57). Mas'üd also unleashed his Hindu troops upon

the rebellious Turkish governor of Lahore, Ahmad b. Niyättigin, or Niälti-gin in 426/1034. Despite

Ahmad's defeat of the first Hindu general sent against him, five years later, in the words of al'Utbi,

Ma'süd sent another Hindu general named Tilak (also recorded in the Tabakät-i Akban bin Jai Sen;

see Elliott, 1867-77b: 110), who was victorious (Nizam, 1993: 400). On this occasion the victory

was owed to a group of Jhäts mobilised by Tilak against the rebellious governor (Elliott, 1867-77b:

II, 54). On another occasion, he sent a force of 500 Indian cavalry and infantry to capture a

rebellious general, whilst he used Indian troops to guard his imprisoned uncle, presumably again

because of their absolute loyalty to the sultan (Bosworth, 1963: 110).

Our most interesting source for information about Hindus in positions of authority in the Ghaznävid

state is the Tarkhu-s Subuktigin of Abii-l Fazl Muhammad ibn al-Baihak (see Elliott, 1867-77b:

III, 53-127), also known as the Tarikh-i Baihaki. As well as the references to campaigns, where the

capture of Hindu populations is described (as in 424 AH/1033 CE when the women and children of

Sarsuti in Kashmir are "carried of by the soldiers as slaves"; Elliott, 1867-77b: III, 56), Baihaki

relates the histories of two senior military commanders, Ariyäruk, the commander of the army of

Hindüstän during the reign of Mahmüd (Elliott, 1867-77b: III, 89-96), and General Tilak the Hindu

(Elliott, 1867-77b: III, 106-112). The fates of the two are very different; one a morality tale of

vanity, drunkenness and eventual disgrace, the other, one of martial success and military glory. The

common point of the two is their depiction in the work as high-ranking Hindus, clearly

demonstrating that there were such. In Ariyäruk' s case, he is portrayed as being powerful enough to

defy the injunction of the sultan; "when he was arraigned in the time of reign of Muhammad, he did

not submit" (Elliott, 1867-77b: III, 89), and as being a drunkard, arrogant and presumptuous. He

had been promoted earlier by Mahmüd (Eliott, 1867-77b: III, 90), and had been one of the partisans

of Mas'üd when the throne had been contested between himself and Muhammad in 1030 CE. In the

tale of the downfall of Ariyaruk, the chief of the guards is mentioned as one "Muhtäj" (Elliott,

1867-77b: III, 94), whilst the capture of the bibulous Indian also involved other Hindu officers and

some three hundred soldiers (Elliott, 1867-77b: 95). Interestingly, the news of the removal of Ariydruk is suppressed, according to al-Baihaki, "so secretly that nobody knew he had been

dismissed", which implies that Mas'üd was concerned about the possible reaction to this arrest

151


(Elliott, 1867-77b: III, 96). The question that arises in this instance is whom the sultän feared -

perhaps the loyalty of the Hindus was not entirely secure.

The second of the generals mentioned by name in al-Baihaki is Tilak the Hindu, in the context of

the rebellion of the Governor of Hindustan, Ahmad Niältigin. Tilak was commissioned with the task

of subduing the disobedient Ahmad, by the Amir Mas'üd Abü Sa'id b. Mahmtid (421-431 AH/1030-

1040 CE) in 427 AH11036 CE, though the other commanders were clearly not happy about this.

Baihaki records a message from Mas'üd to Tilak, in which he expresses his perceptions of the

situation; "You have shamed them all... You shall be raised to a higher rank... They are greatly

annoyed at your exaltation" (Elliott, 1867-77b: III, 107). This suggests that the Hindu general might

have been one of those who had initially supported Muhammad in the contest for the throne against

Mas'üd during the period immediately following their father's death in 421 AH/1030 CE, and not

part of the intimate circle of "boon companions" who would be "greatly annoyed", but clearly

Mas'üd tied the Hindu officers to him through patronage and favour as a balance to the Turkish

"boon companions". The text does not refer to the earlier expedition sent by Mas'üd that was

defeated (Nazim, 1993: 424), but it does refer to Tilak as studying in Kashmir, where he

"... acquired some proficiency in dissimulation, amours, and witchcraft", the latter an unusual skill

for a general perhaps, and his role as interpreter between the vizier and the Hindus, enabling him to

gain great influence at the minister's court (Elliott, 867-77b: III, 107-8). In this capacity he would

be one of the necessary conduits for communicating with the troops, and perhaps one of the

architects, if we could describe them as such, of the koine that would eventually become Romanes.

He would appear to have been noticed in this service by Sultän Mahmüd himself, although when

this took place is not recorded. His service was such that he was able to persuade all of the Hindu

"... and many outsiders under his rule", by which al-Baihaki appears to refer to a body of troops

organised as part of the Ghazndvid armies under Mahmüd (see Elliott, 1867-77b: A, 403-404, for

his conjectures upon the origins of these amongst the Central Asian Chi, and their relationship to the

Jhäts). Tilak is elevated at the expense of another Hindu general, Sundar, just before his campaign

against Ahmad Nldltigin, and here may be the reference to the earlier defeat, as "Sundar, the general

of the Hind, was not in his place. " (Elliott, 1867-77b: III, 109) As mentioned above, Tilak was able

to accomplish his mission thanks to being able to mobilise numbers of Jhdts against the rebel and

utilising their skills as sailors.

The Ghaznävid army was frequently commanded by Indians, made up of many contingents and Hindus, and attended on campaigns by Jhäts and other Indian auxiliaries, some of them tributary

152


troops (such as those received from the ruler of Hind, two thousand men annually; see Elliott, 1867-

77b: D, 444). The numbers of troops were substantial and the attendant community in the capital, in

its own mahalle, correspondingly large. Even after the defeat at Danddngan, the Amir Mas'üd

retained many Indians in the Ghaznävid military machine, coming to rely upon them more heavily

than other groups for support against rebels and those dissatisfied with his continued rule. This book

argues, in the context of the earlier discussion (Hancock, 2006; Matras, Hübschmannovä, 2004,

Fraser, 1992), that these Indians, in both their military capacity and their auxiliary or domestic roles,

are the ancestors of the Romani peoples, and that furthermore, the current debate about the military

or non-military origins of the Gypsies is misplaced, if it is seen to rely upon a single causal factor,

language. As Thapar has argued in relation to the Aryans, "[t]he historically relevant question [is]

not the identity of the Aryans (identities are never permanent), but why and how languages and

cultures change in a given area... "(Thapar, 2000: 15).

My suggestion that Gypsies had a particular status as a taxable community in the complex ethno-

religious mosaic of the Ottoman Empire, related to a series of economic (if we count war as the

acquisition of wealth by another means, to paraphrase von Clausewitz, 1908) and social niches,

engages directly with this debate. The process whereby this was fundamentally undermined was the

intrusion of western European orientalist imperialism. This dramatic deterioration eroded a set of

relations that had existed since the earliest times in the east under Islam, and Oriental Christianity

for centuries, and that was not reflected in social and economic relations western Europe. The

conditions that nurtured the emergence and development of the Romani peoples were set in place

during the Ghaznävid period. The mechanisms and processes of this emergence are those that I have

attempted to outline here, to challenge the imprecision and inaccuracies of the previous histories,

and to respond to the studies of Hancock, Matras, Fraser and Hübschmannovä in some meaningful

way. The detailed articulation of the Ghaznävid system is a means of illustrating the processes to

establish the patterns that have underpinned the Romani experience under Islam and Oriental

Christianity (Armenian and Byzantine Orthodoxy, in this case), as the model for the Ottoman

picture prior to its change. The place of Hindus as warriors, auxiliaries, entertainers, slaves and

servants of the sultan in this earlier phase profoundly influenced the social structures both of the

proto-Romani and later Romani groups in the East, and the treatment of them by those states in

which they found refuge (as opposed to Western Europe where refuge from enslavement turned

rapidly into repression). A detailed understanding of the degree to which Indians were integrated

into Ghaznä, even to the extent of being part-and-parcel of the razzias against their quondam

153


homeland, exemplifies the complexity of the debate about origins, military and non-military, and

the importance of seeing the period under Muslim rule for these particular groups (c. 997-1040 CE),

as the "cradle" of the Romani people.

The nature of the campaigns that Hindü commanders led, and Indian troops (see the illustration

above of the Ghazndvid forces besieging a city from an MS of Rashid al-Din's Djämi al-Tawan-kh

in Edinburgh University Library; Stiliman, 1986: 21-2), were engaged in will be discussed in more

detail in the next section, but here it is important to note that there are clear distinctions highlighted

by the British historians of the 19th century, in their portrayal of Hindus in these different contexts.

These interpretations had a profound impact upon how these events were viewed subsequently, and

depended upon an interpretation of the campaigns as exclusively Muslim conquerors pitched

against tenacious, but ultimately unworthy opponents, the Hindus. Elliott condemns the ignorance

of the latter in relating the circumstances surrounding the first campaign of Sebüktigin near the

Lamghän valley, when he suggests that a snow-storm had confused the Indians and they had

suffered badly in the ensuing cold and frost. The explanation for this in the Kitäbu-1 Yämini of al-

'Utbi and the Jawämi'u-l Hikäydt wa Lawä-mi'u-l Riwäyät (c. 607 AH/1211 CE) of Muhammad 'Ufi

at Delhi (see Elliott, 1867-77b: IV), a "Collection of Stories and Illustrations of Histories", is that a

miraculous spring in the vicinity emitted thunder, storms and freezing weather when the "Yedeh"

stone was thrown into the water, and the "minds of the natives of India would naturally have tried to

account for such a supernatural phenomenon... and superstition was at hand to render her

assistance. " (Elliott, 1867-77b: D, 435) The magical stone however, derives from a Turkic legend

that was still current amongst the Mongols in the thirteenth-century, that Marco Polo records in his

The Book of Ser Marco Polo the Venetian concerning the Kingdoms and Marvels of the East (Yule,

1903: 206), yet Elliott clearly sees the "superstitious" Hindus as in a different category to those like

Tilak the General in the Muslim armies of Amir Sebüktigin, Sultan Mahmüd and Amir Mas'üd

(Elliott, 1867-77b: D, 435). The actual nature of these campaigns, like the armies that undertook

them, is far more complex than this simple dichotomy allows.

154


4.6 The Ghaznävid campaigns in India: ghazawät and razzias

THE NARRATIVE SOURCES RECORDING THE ASSAULTS on the region of northern and central India

following the assumption of power of Amir Sebüktigin in 367 AH/977 CE (see Briggs, 1829: 11-

27), and later Sultan Mahmüd of Ghaznä (387 AH1997 CE), provides us with some descriptions of

Indian peoples in the empire as captives, slaves, commanders, generals, cavalry, infantry, elephant-

drivers, prisoners of war, craftspeople and artisans. They also celebrate and glorify the victory of

Islam over the heretic and infidel, and exaggerate the weight of booty, numbers of slaves and

destruction of cities that the Sultän inflicted upon his enemies, as Thapar has argued (2000a: 31-3).

The incorporation of these texts into the British colonial discourse provided some justification for

the imperialism and the expansion of the Empire, and it is interesting to see the shifts in rhetoric

that accompanied this project, the collusion of scholarship in the colonial enterprise. Elliott's

editions of the sources do not decry the "Hindu trauma" in the way that the earlier Governor

General, Lord Ellenborough had done in the 1840's. They promote the opposite view, in that the

Hindu had been subjugated by the might of the `ferocious' Mahmiid and that his rule paved the way

for the British, almost that the Hindu peoples deservedly were ruled by others, not being fit to do so

themselves, according to Elliott (1867-77: 3). This picture of irreconcilable conflict and endemic

ancient hatred was part of the discourse that it was believed, justified the British colonial

administration and rule over India - as peace-keepers protecting the interests of the merchants and

ultimately the Crown, but by the time the Elliott editions had been printed in the early 1880's, the

logic of imperialism and the post-Darwinian notions of progress, advancement and civilisation had

all come to `ennoble' the `White Man's Burden' as Kipling would come to call it (though initially a

poem for Queen Victoria's Diamond Jubilee, it was subsequently addressed to the American's after

the conquest of the Philippines; 1899). The manipulation of this discourse was itself an aspect of British domestic party politics (Thapar, 2000a: 44-45), with the `usable past' being deployed to

ultimately ensure the profits from trade remained at their maximum (see Spiegel, 1993 for notions

of "romancing the past"). The fostering of an antagonistic duality of Muslim-Hindu relations,

through scholarship such as Elliott and Dowson's History of India... and their interpretations

thereof, was integral to the colonial project. The campaigns carried out by Mahmüd were driven by

the same economic forces and justified by the same sort of rhetoric.

The sources do record the capture of Hindus by Mahmüd's forces and the necessity of this element of Ghaznävid razzias both as a source of human labour and fresh troops, and for the commercial value of the slaves (Levi, 2002: 281). The campaigns agianst the Hindus were expensive, with

155


months in the field and the support of large numbers of troops, auxiliaries and animals draining the

treasury. The primacy of recouping those costs and returning a profit on each venture is clear in the

chronicles; Firi*tah frequently refers to the amount of treasure, the numbers of slaves and animals

captured. In Bhatea, during the campaign against Beejy Ray (1004 CE/395 AH), Mahmüd

succesfully defeated the prince and took the city, capturing "Two hundred and eighty elephants,

many slaves, and other spoils, obtained in the town"(Briggs, 1829: 28). Even when the motive for

war was purportedly religious, the financial rewards were always close behind such considerations:

"The King, in his zeal to propagate the faith, now marched against the Hindoos of

Nagrakote, breaking down their idols and razing their temples. The fort, at that time

denominated the Fort of Bheem, was closely invested by the Mahomedans, who had first

laid waste the country around it with fire and sword. Bheem was built by a prince of the

same name, on the top of a steep mountain, where the Hindoos, on account of its strength,

had deposited the wealth consecrated to their idols by all the neighbouring kingdoms; so that

in this fort there is supposed to have been a greater quantity of gold, silver, precious stones,

and pearls, than was ever collected in the royal treasury of any prince on earth... " (Briggs,

1829: 29)

Indeed the plunder in this period was such that Firishta laconically remarks that recording "the

details of which would be tedious... " (Briggs, 1829: 29).

The incorporation of large numbers of slaves into the Ghaznävid state was a consequence of this

need for warriors, artisans and labourers, in the same way that other non-Muslim peoples had been

coerced or co-opted during the period of the early Islamic conquest of Sindh (93 AHl712 CE; see

Gibb, 1932). These ghazawät or razzias were a continual aspect of their policy of campaigning in

the sub-continent against Shi'ite Multan and other centres of the Isma'ili Muslims of Sindh, and the

infidel Hindus. Sources suggest that this policy was first embarked upon by Amir Sebüktigin, whilst

he was still under the command of Alptigin, the Sämänid mamlük who seized control of Ghanzä

after 351 AH/962 CE (Bosworth, 1991: 66). Firishta writes that "Alptigin... retained his

independence fifteen years, during which period his general, Su-booktugeen, being engaged in

frequent wars with the Indians, as often defeated them (A. H. 365/A. D. 975)" (Brigg, 1829: 11), but

the dates of these earlier incursions are not clear. al-`Utbi records that the Amir Sebüktigin "... made

frequent expeditions into Hind, in the prosecution of Holy Wars" (Elliott, 1867-77b: 16), and

suggests (incorrectly) that these were the first incursions by Muslims, when he states that the "cities

in Hind, ... had up to that time been tenanted only by infidels, and not trodden by the camels and

156


horses of the Musulmäns" (Elliott, 1867-77b: 16). Firishta suggests that Amir Sebüktigin made

several raids upon Hind, in addition to those he described in more detail at Lamghän and Kusdär

(see below). From the very first, it would appear that Mahmüd had an important role in these, albeit

by proxy, as his father celebrated his birth by destroying the idols of Sodra on the Chinäb river, in

361 AH/971 CE (Elliott, 1867-77b: Note D, 2; Bosworth,, 1991: 65). Al-`Utbi describes the assault

and plunder of Lamghän in 369 AH/988 CE, in pursuit of the Raja Djaipal (Jaipäl), whom he

eventually defeated (Longworth-Danes, 1993: 155). Djaipal's resistance to the Islamic incursions

earned further punitive raids in 378 AH/988 CE, this time with both the Amir and his son Mahmüd,

whom al-`Utbi describes as "... like a lion of the forest or a destructive eagle... " (Elliott, 1867-

77b: 17). This zeal was the cause of conflict between the two, as Sebüktigin appears to have

censured his son's refusal to negotiate with Djaipäl's envoys by a short imprisonment afterwards,

for presumably challenging his father (Longworth-Danes, 1993: 155). This may have been behind

the unusual decision of Sebüktigin to appoint his younger son Ismäil as heir, upon his death in 387

AH/997 CE (Haig, 1993: 133). Mahmüd defeated his brother and assumed the title of "Amir" in

389 AH/997 CE, also being awarded the laqab (honorific) of Yam-in al-dawla Amin al-milla (hand

of the Khalif and defender of the faithful) by the Khalifa al-Kadir, one year later (Cole, 2003; Haig,

1993: 133; Longworth-Danes, 1993: 155). A pattern begins to emerge in the reign of Amir

Sebüktigin that is followed by his son with enormous success, and that sets the seal upon the

dynasty and its orientation, as a whole. The wealth of India is suborned by the sultans to meet the

needs of the Ghaznävid state, under the banner of legitimate ghäzwarä against the infidel Hindus.

Mahmüd would appear to have conducted almost yearly cold-weather campaigns into India during

the entire period of his rule, although the exact number is unclear (Longworth-Danes, 1993: 155).

Elliott notes that Nizämud-din Ahmad's Tabakät-i Akbari mentions twelve but actually records

sixteen, whilst Dow's 1792 translation of Firishta notes twelve initially, but enumerates fifteen

(Elliott, 1867-77b: D, 1). Elliott himself details some seventeen (1867-77: D), and a more modern

historical examination by Haig (1913-1936) agrees with this figure (1993: 133-135). The difference

in number relates both to the lack of sources, and the confusion over what constituted individual

raids and at what times they took place. As Elliott notes in his introduction to "Mahmüd's

Expeditions to India" (1867-77: D, 434-466)

The times places, and numbers of Mahmüd's expeditions to India have offered great

difficulties to those who have dealt with the history of that ferocious and insatiable

conqueror. We look in vain for any enquiry on the subject from the native historians of this

157


period, who, in their ignorance of Upper India, enter names and years without the scruples

and hesitations which a better knowledge or a more critical spirit, would have induced... "

(1867-77: 434)

In spite of the obvious prejudice of this statement (the record of 11th century Britain is hardly a less

confusing record; see Insley, 2000), the observation of the paucity of contemporary non-Muslim

records is initially striking, suggesting a calamity too terrible to be remarked upon by those whom

the Ghaznävids despoiled. Yet this image of the absence of a Hindii perspective upon the period

may, in fact, tell us more about the justification for British rule than about the lack of contemporary

Hindu historians of Indian history. In his obituary for the noted Hindu historian and scholar

Bishambhar Nath Pande (1906-1988), F. M. Faruqhi reiterates the question asked by Elliott

regarding the paucity of Hindu sources, and suggests that "Sir Henry, and scores of other Empire

scholars, went on to produce a synthetic Hindu versus Muslim history of India" (1998). The

distortion of Indian history, suggested Pande, served the interests of the British colonial governors

in both dividing the population of India and in neutralising antagonism towards themselves by re-

directing hostility towards each other. Faruqhi quotes Lord Elgin as Viceroy of India (1862-63),

cynically remarking that "We have maintained our power in India by playing-off one part against

the other and we must continue to do so... " (1998: 22). In this context, Elliott's History... is an

exercise in colonial rule, rather than a history conducted "with better knowledge or a more critical

spirit... " (1867-77: D, 434). The wider discussion of bias and misinterpretation in Indian history

writing (and its promulgation in "nationalist" textbooks under the BJP government; see Deb, 2003),

have been discussed above, but the implications for Romani historiography, particularly the re-

writing of Romani history, are profound. The centrality of Mahmüd of Ghaznä's razzias to the

Indian ethnogenesis theory are well-established, and almost universally accepted (outside of the

"Dutch school"; see Mayall, 2004: 220-29), as the major factor in the migration of the Gypsies from

India (as opposed to the ancestors of groups who coalesced and emerged as a distinct Gypsy

identity in Anatolia, the position adopted in this book). If, as seems clear, the picture of Muslim

incursions, and enslavement of huge numbers of Indians needs to be qualified by the perspective of Pande, Faruqhi and others, how are we then to account for the presence of these Hindus in Ghaznä

and thus provide a plausible context for the "origins" of the Romani peoples? The answer to this

may lie in the trope in Romani historiography of the Gypsy as victim, as pariah and outcast. As

suggested earlier in this work, the political importance of claiming such a status has to a large extent determined the kind of history produced in support of such claims. This perspective would require

158


that the understanding of the social and economic position of Indians in Ghaznä would be a

consequence of their capture and enslavement. As I shall describe below, this image is challenged

by the evidence we have of Hindü participation on a broader scale and their positions of authority

that they occupied in the structures of the state and army. It is also a picture that is problematic in

view of Mahmüd's own behaviour, as suggested above; his sponsorship of heterodox Muslim

groups in Khorasan and his wholesale incorporation of artefacts and motifs from India in the

construction of the most important mosque in Ghaznä, the Arüs al-Falak, or "Celestial Bride of

Heaven" are not entirely consistent with the ghazi image, nor the Hindutva picture of the rapacious,

tyrannical Hindü-killer.

The interests of those writing for the sultan about the razzias were clearly to celebrate the

superiority of Islamic arms against the infidel, and al-Ùtbi is fulsome in his praise of Muslim arms,

as when Amir Sebüktigin invades Hind in 369 AH/979 CE. During the Lamghan campaign, "he...

captured other cities and killed the polluted wretches, destroying the idolatrous and gratifying the

Musalmdns... wounding and killng beyond measure... " (1876-77,11: 18). Whilst Bosworth notes

that his jal-Ùtbi] "ornate and verbose style" does not ignore the darker side of the sultans' reigns

(2000: 945), his confusion of dates and lack of knowledge about the territory concerned makes him

less reliable (Elliott, 1867-77: C, 429). His depiction of the "darker side" of these episodes is not

unambiguous, but it is hard not to see these descriptions as merely triumphalist. Elliott echoes al-

Ùtbi's perspective of the accounts of Hindü defeat, when he comments on the ease with which

"... the hardy warriors of Zabulistän had overcome the more effeminate sons of India" (1867-77: D,

434). As Lal has pointed out, the contrast for the British between the colonial "administration"

(rather than "rule") and its subjects, is frequently portrayed in terms of "vigorous" and "improving"

government, as opposed to "ancient" and "religious" (read "degraded" and "effeminate") society

(Lal, 1993). Clearly the imperial project of the British sought antecedents in the Muslim conquest of

India eight centuries earlier. Conversely, the early nationalist struggle chose to ignore this

dichotomy in favour of portraying a pre-colonial society where,

"[w]e have chapter and verse given to us by Hindü historians and by Mussalman [sic]

historians to say that we were living in comparative peace even then. And Hindus and Mussalmans in the villages are not even today quarrelling. In those days they were not known to quarrel at all. " (Gandhi, 1931)

In this light, the campaigns of the Ghanävids were to be subsumed under the notion of a "comparative peace", in the interests of opposing those who sought to claim their spirit as

159


"ferocious and insatiable" conquerors, masquerading behind a discourse of advancement and

improvement, the British. The actual nature of the campaigns themselves is difficult to ascertain in

the writings of the Muslim chroniclers, filled as they are with the assumptions of superiority and

"the subversion of idolatry, so shall their reward be in heaven" (Firishta, quoted in Elliott, 1867-77:

D, 448).

The Indian troops in Mahmüd's army engaged in the Indian raids made up a sizeable element of the

infantry forces, the piyädagän-i dargähi, mounted on camels for long distance campaigns, pitched

battles and sieges (Bosworth, 1963: 114). These slave corps were commanded by the Salär-i

Ghulämän, an officer second only to the commander-in-chief, usually but not always a Turk

(Bosworth, 1963: 99). These were armed with a variety of weaponry, some of it specialised (Indians

used the lasso, for example), and produced in India, such as steel cuirasses and sword blades

(Bosworth, 1963: 120-1). In addition to these troops, locally recruited soldiers also played a part in

the campaigns, together with auxiliaries who were blacksmiths, carpenters, stone-breakers and siege

engineers. As suggested earlier, the army also brought along supernumerary personnel such as

astrologers and large numbers of merchants in the bäzär-i lashkar or `army market' (Boswortth,

1963: 119). One of the major logistical problems for an army on the march was the control and

management of these additional (and not always welcome) civilians (Hale, 1989: 63).

The accounts of the assaults on India provide some idea of the scale of the plundering of temples,

cities and the country. For example, after narrowly defeating a combined force of Gukkars

(Ghürkäs) and "the Rajas of Ujjain, Gwäliär, Kälinjar, Kanauj, Dehlf, and Ajmir" (Brigg, 1829:

1,32) at Und in 399 AHJ1008 CE, Mahmüd plundered the wealth of the temple of Nagarkot, or Kot

Kängrä (Haig, 1993: 133). Additional treasure was gained by the fact that the "Hindoo females, on

this occasion, sold their jewels, and melted down their golden ornaments (which they sent from

distant parts), to furnish resources for the war" (Brigg, 1829: 1,32). Firishta notes the extent of the

temple's treasures as

... 700,00 golden dinars, 700 muns of gold and silver plate, 200 muns of pure gold in ingots,

2000 muns of silver bullion, and twenty muns of various jewels, including pearls, corals, diamonds, and rubies... (14001b. of gold and silver plate, 4001b. of golden ingots, 40001b. of

silver bullion, and 401b. weight of pearls, corals, diamonds, and rubies) (Brigg, 1829: 33)

The fabulous treasure secured from the temple at Somanath in the campaign of 415 A14/1024 CE,

was seized with the troops and large numbers of irregulars, such as the "... 30,000 youths of

160


Toorkistan and the neighbouring countries, who followed him [Mahmüd] without pay... " (Briggs,

1829: I, 42). This element was a constant factor in the Ghaznävid forces, and no doubt influenced

the choice of missions undertaken for plunder. The army under Mahmüd was an enormously

effective institution, comprising of a differing elements of professional soldiers, elite palace guards

and a militia from the tribal Türkoman who assembled at the start of each raid, for plunder and

booty. The auxiliaries were augmented by local forces of various Indian origins, and allies, as well

as ethnic units of the kind that had been present in Muslim armies from the earliest times (see

Crone, 2003: 287-285; 1991: 877). In these complex elements we may find the origin of the

Romani people.

161


4.7 The city as theatre: Ghaznävid culture & literature

THE EXPANSION OF GHAZNA DURING MAHMCD'S RULE, especially following the important razzias of

408-409 AH/1018-1019 CE against the cities of the northern Ganges, Kanuaj and Mathura (see

Elliott, 1867-77b: D, 450-455, for a summary of the sources and a discussion of the probable route),

was financed by the immensely valuable treasures of India. Looted and carried by captive Hindus,

this wealth supported not only the military, but the cultural and literary growth of the Ghaznävid

state. Historians of the period, such as Firishta, record the splendour of palaces and public buildings

endowed by the Ghaznävid nobility, after Mahmüd's construction of his great central mosque, the

"Celestial Bride of Heaven". The complex also included a university "... supplied with a vast

collection of curious books in various languages... [and]... a museum of natural curiosities"

(Briggs, 1829: 1,39). In order to support this educational endowment, Mahmüd is described as

bringing what Firishta describes as the "learned men who lived at the court of Mahmood... "

(Briggs, 1829: 50-1). The Shah-näma of Firdaws belongs properly to this period (c. 401 AH/1010

CE), when scholars, poets and philosophers were brought to Ghaznd after Mahmüd's conquest of

Khwarzmshah in 408-409 AH/1016-1017 CE, though not always willingly. The historian and

scholar, al-Birüni, the mathematician Abü Nasr Mansur and the physician, Abul-Khays were all

members of this group. Firishta writes of the four hundred poets and learned men, besides the

university students at Ghaznä (Briggs, 1829: 1,51). The festivities surrounding this erudite

company, including the majälis-i nashät u sharäb, or wine-drinking sessions, often featured

recitations from poets and scholars, as well as the clowns, musicians and dancers (including

Indians), that accompanied such events at the palaces (Bosworth, 1963: 138). Bosworth suggests

that the mosque complex, and the palaces and other buildings in the city, bore clear influence of

Indian craftsmanship, and incorporated objects made from precious metals and Hindu statuary in

their fabric, captured as war trophies (1965: 1049), a surprising feature in an orthodox Sunni

environment. The ruins of the lashkari bäzär, or army market, still display frescoes of rows of

palace ghuläms surrounding the sultan's throne in the principal audience chamber, amongst whom

appear images of the Ghaznävid Hindus (Sourdel-Thomine, 1983: 1054). Again this suggests a

more complex character than the rapacious and fanatical Muslim war-leader portrayed in test

referred to earlier. The construction of towers of victory, or more possibly minarets of very

particular design and execution, in the later Ghaznävid period under Mas'iid III (1099-1115 CE)

and Bahräm Shall (1118-cl157 CE), reflect the impact of the skilled craftsmen and artisans of India

"... as Zäbülistän had no artistic traditions of its own. " (Bosworth, 1983: 1049)

162


The military expeditions to India undertaken after this conquest were accompanied by al-Birüni,

who learnt Sanskrit and some of the dialects to produce his Description of India or Tarikh al-Hind

(421 AH/1030 CE), a remarkably balanced work for its time. Unusually, the cycle of stories relating

the exploits of Bahräm Gür (Shäh Varakhan V) contained in Firdawsi's Shäh-näma, and most

especially those relating to the "injustices of the King of Hind" may have reflected not only the

Ghaznävid perception of Indians as enemies, but their complexity of views towards Hindu

Brahmans and Rajas. More importantly perhaps, the cycle contains the interesting episodes of some

length, about the relations between the Shah and Shangul, the king of Hind (VII, 109-148),

including of course, the episode of the Gypsies. In the context of the genre of "mirrors for princes"

such as the Qäbtzs-näma, it might be argued that Mahmüd consciously strove to imitate some

qualities of the Iranian king (Bruijn, 1987). The image of the Ghaznavid state was clearly meant to

rest upon notions of culture and learning, not merely the acquisition of wealth and plunder, or the

military success of Islam. The later award of titles to Mahmüd by the Khalif at Baghdad in 417

AH/1027 ("Guardian of the State and of the Faith"; see Briggs, 1829: 47), and to his sons, the

Amirs Mas'üd and Yüsüf, reflect the esteem the Ghaznävid court was held in by the wider Islamic

world, notwithstanding the sometimes difficult relations with the Khalif (Bosworth, 1991: 66). This

court was to be celebrated not just in its heyday, but by the Ghaznävid's successors, the Saldjtikids

through the work of Nizam al-Mulk, their great vizier, who would refer to them as a model of

efficient rule and organisation in his Siyäsat-näma or Siyar al-mulük, the celebrated Book of

Government or Rules for Kings (Darke, 1960). It is to this state that superseded Mahmüd and his

heirs in the western Muslim lands, that the next part of the history of the Gypsies belongs.

163


5.1 The Saldjükids, Armenia and Byzantium; "the divine-rebuking wrath of God... the savage

nation of infidels called Turks... "

IN ORDER TO UNDERSTAND THE CONTEXT of the arrival of what can be described as a proto-Romani

speaking, (Hancock, 2005: 25; Matras, 2004: 278), or Rajputic (Hancock, 2000: 23) peoples in the

lands of Byzantium, it is necessary to briefly examine the `origins' of the Saldjük, Selcük or Saljuq

Turks (referred to above by Matthew of Edessa, describing their first appearance in the Armenian

lands in 467 AE/1018 CE; see Dostourian, 1993: 47). This princely Turkish clan were descended

from the Oghuz Türks (see Cahen, 2001: 21-65; 1968,19-119; Luther, 2001; Bosworth, 1986: 936-

959; Houtsma, 1993: 208-213; Talbot Rice, 1961), or Ghuzz Turks as they are known in the Arab

sources (see Holt, 1983: 1106-1111). In the context of 11th and 12th century Asia Minor/Anatolia, it

is also important to reflect upon the relationship between the Saldjüks, the Turkmen or Turcomans,

and the Ghaznävids (Houtsma, 1993; Langdon, 1992; Vryonis, 1971; Barthold, 1945). This is in

order to establish a clear background for the primary intention in this chapter, viz, render a plausible

trajectory of movement for the proto-Romani from Ghaznä to Constantinople, and a time-scale that,

although rapid, confidently pursues the arguments referred to and established previously in this

book. Hancock has most recently summarised the debate and his earlier arguments in what he has

called "the Seljuq factor" (2006: 83-84), in opposition to scholars who have disputed the speed with

which this population is likely to have moved (Hübschmannovä, 2004; Iovild and Schurr, 2004:

275). This book additionally makes the argument for a rapid dispersal of proto-Romani into

Anatolia, and attempts to elaborate on the likely historical context of such a migration, including the

role of the Turkmen, led by the Saldjüks, as the significant factor in the migration of these groups.

Here, I will also investigate the possibility of these proto-Romani being involved as mercenaries, in

the defence of early mediaeval Armenia, before the kingdom's destruction by the combined intrigue

of the Byzantines, and military defeat by the Saldjükids in 1064 CE.

The role of the Saldjüks in the Turkification of Byzantine Asian Minor is also referred to, as it has

been the source of some heated debate, especially amongst nationalist Turkish historians (see

Bosworth, 1995: 937), who see the Saldjük empire as the primary Muslim Turkish power to

establish itself in the central lands of Islam. This perspective, still at the heart of Turkish teaching

and textbook histories today, treats the Ghaznävid and Qaraq . nid polities as peripheral, in favour of

a lineage stemming from the sons of Saldjük b. Lugmän to the Republic of the present (see

Kafesoglu, 1988; Bosworth, 1999). The Turkification of the Byzantine lands has also been

duscussed by Vryonis extensively in his monograph (1971), a process that he suggests was

164


unusually long, as a "... drawn out period [that] subjected Byzantine society to repeated shocks and

dislocations... in a continuous process that lasted four centuries. " (1971: 143) The loss of Asia

Minor to Byzantium, the destruction of the mediaeval Bagratid Kingdom of Armenia and the

establishment of the Sladjükid Sultanate of Rüm, as an appanage of the Great Saldjüks of Persia,

can all be traced to the defeat of the Ghaznävids on the field of Danddngan (431 AH/1040 CE), as

the Ghaznävid Empire had acted as a bulwark against the incursions from the Turkmen from

Central Asia.

This crucial period in Romani history is the most difficult, in that there are literally no sources that

can be cited, referring to Hindus, Indians or any group that one might suggest may be proto-Romani

in any guise, apart from one intriguing reference in the chronicle of Matthew of Edessa to the

numbers of refugees and destitute people flowing through Asia Minor in the mid-to-late eleventh

century (Dostourian, 1993: 144). In this context, the historian's task is most akin to White's

suggestion of an exercise in "constructing" a narrative from fragments of the past that remain to us,

and thus can be challenged as producing the most speculative and conjectural of "histories" (White,

1987b: 8). Acknowledging this to be the case, nevertheless, what is presented here is an attempt to

explain a suggested relationship between the remnants of the Hindu soldiery at Dandängän in May

1040 CE, and the appearance of groups of people defined as Egyptians or Aigupta (Aigupta) in the

Byzantine capital of Constantinople in the late eleventh-century. In order to "reconstruct" (or

"construct") such a relationship, it is inevitable that a "narrative of journey" is established (which I

have described as the dominant trope of Romani Studies scholarship). This, to use the terminology

of White, is the "poetic mode" in which the discourse is framed, and "tragedy" becomes the major

device for "emplotment" of the narrative (see White, 1975: xi). In a sense, this exemplifies White's

discussion of the Chronicle of St Gall. The Chronicle proceeds to record a series of events in the

form of an annotated list of dates, such as "709 (CE), hard winter, Duke Gotfreid died", "710 hard

winter and deficient in crops", and "712 flood everywhere" (White, 1987b: 6-20). The point of

reference here is not the "events" listed, but the years when nothing is recorded at all; 711,713,

715,716 and 717. The absence of any narrative content is taken by White to be an illustration of

how "... social events are apparently as incomprehensible as natural events. They seem to have the

same order of importance or unimportance. " (1987b: 7). As White goes on to argue, "... narrative

strains for the effect of having filled in all the gaps, of having put an image of continuity, coherency,

and meaning in place of the fantasies of emptiness... " (1987b: 11) The exercise here is to "fill in all

the gaps", having acknowledged that to desire to do so, is itself an attempt to give meaning to a

165


series of events, recorded without mention of groups I can suggest are connected to my over-

arching narrative, to what might be described as years when nothing is recorded at all.

The trajectory, or narrative of journey outlined here is then, a construct, an alternative vector of

coherency and continuity, and it must be recognised as such in order to avoid being judged as if it

were indeed a "real" series of events being described, and subsequently argued over in opposition to

alternative, and equally conjectural, trajectories. The very structure of the debate around the "Seljuq

factor", as Hancock typifies it, is one that assumes it is possible to connect the history of the

Gypsies in a meaningful way with that of the Saldjüks, to demonstrate the place of one group in

relation to the other. To date, no secure information has been drawn from sources (narrative or

documentary), to allow us to infer the kind of positions currently being adopted, nor to refute those

of others. All these must be measured as providing more or less informed conjectures about the

period in question, and consequently their likelihood as sufficient attempts to address the problems

posed by scholars. As Acton has remarked, "with sociologically informed questions prompted by

archival research, we end up with a greater capacity to discriminate between speculations... "

(Acton, 2004: pc) The multiple questions raised by scholars might be summarised as can we find

enough information, in the context of the turmoil and chaos of mediaeval Anatolia, to suggest a

connection between those identified groups of Egyptians that appear in Byzantium in the late 11th

century with those we have identified in Trans-Oxiana in 1040 CE? The underlying problems of

whom we are defining, and in what manner these definitions have been assumed, must also be

carefully considered. It must also be acknowledged that the mode, form and emplotment used have

been selected a priori, though as Partner has illustrated, the point at which this has taken place is

notoriously difficult to pinpoint (1997: 108-109), and is almost inherent in the approach to the work

in its infancy, not a conscious decision once the material has been gathered, prior to its being

ordered, rationalised and narrated. As I have outlined it above, this narration moves from a brief

consideration of origins of the particular group of Oghuz we know as Saldjük, to a consideration of

the circumstances recorded by the chronicles (some contemporary, some not) written about the

region in question. By this means, I hope to illuminate salient points that will allow me to challenge

some aspects of the "Seljuq factor" arguments from scholars engaged in making or refuting them,

and propose a likely scenario for a relationship existing between Ghaznavid Hindus and Byzantine

Egyptians.

5.2 "... destiny decided that he should live a reign of pain and vexation... " Mas'üd b. Mahmiid and the fragmentation of Ghaznävid authority

166


THE CAMPAIGN THAT LED TO THE DEFEAT OF MAS'ÜD'S ARMY in 431 AH/1040 CE at Dandängän was

itself troubled by numerous problems, with roots in the very earliest days of the reign of Amir

Mas' üd. The initial power struggle between the Amir and his brother Abü Ahmad Muammad, had

left a legacy of division and instability in the Ghaznävid state, after 421 AH/1030 CE. Those who

had remained loyal to Muhammad, including many of the Hindü troops, were at first treated fairly

by Mas'üd after his deposition of his brother, but subsequent executions and imprisonments of

elements of the elite undermined this. Muhammad 'Ufi relates a story of a conversation between

Abü Nasr Miýkdn and Sultan Mahmüd, as a device to explain the fall of the House of Ghaznä:

... Mas'üd, " he continued, "is a proud fellow and thinks there is nobody better than himself.

Muhammad is stout of heart, generous, and fearless, and if Mas'üd indulges in pleasure,

wine, and the like, Muhammad outdoes him. He has no control over himself, has no

apprehension of Mas'üd, and is heedless of the important concerns of life. I fear I find but

little satisfaction in the thought of Muhammad succeeding me; for woe to him at the hands

of Mas'üd, who will devour him, and woe also to the generals of my army, for Mas'üd is a

very covetous man and has great love of money. If he should hear of any officer possessing

a little property, he will be sure to destroy him in a few days, and appoint some worthless

fellow in his place. " (Mohamed Khan, 1835: vol. 2, section 45)

Mas'üd chose to insist on his position by the frequent use of force, but unlike the Sultan Mahümd,

this came to be seen as tyrannical and arbitrary, primarily because the Amir's efforts were seen to be

acting without God's approval. Baihaki emphasises this aspect of the understanding of Mas'üd's

failure when he writes,

Let wise men reflect upon this... that man by mere labour and exertion, notwithstanding that

he has property, armies, and military stores, can succeed in nothing without the aid of

Almighty God. In what was Amir Mas'üd deficient in all the appurtenances of a king?

Pomp, servants, officers of State, lords of the sword and pen, countless armies, elephants

and camels in abundance, an overflowing treasury, were all his, but destiny decided that he

should live a reign of pain and vexation, and that Khoräsän, Khwärizm, Rayy and that

Jabbäl should depart from his hands... "man has no power to strive against fate"... (Briggs,

1829: 127)

The power-play between differing groups within the highly militarised state structure was

exacerbated by the Amir, as the example of Tilak the Hindü (quoted earlier) illustrates. Mas'üd

167


sought support from those who would help him challenge the authority of his father's generals and

viziers, and in consequence, he came to rely heavily upon the Indians, further alienating the Turkish

sübaýi and ghuläms. This was to prove fatal to the outcome of the battle in 421 AH/ 1040 CE, as

many of these high-ranking soldiers (and their retinues) were to become renegades by changing

sides on the battle-field itself. Mas'üd's eastern orientation, together with his failure to secure the

support of his father's lieutenants and integrate his own commanders into the power structure left

him consistently vulnerable to rebellion and dissent, as the frequent challenges to his reign show.

Unlike his father's deposition of his brother Ismä'il in 997 CE, Mas'üd was not universally

acknowledged as the legitimate claimant in his contest with Muhammad in 1030 CE. The passage

above refrains from recounting the previous acclamation of Mas'ud as the heir to the throne in 406

AH/1015-16 CE, and his subsequent appointment as governor of Herat in 408 AH (Nazim, 1993:

400). Although at one time he had been openly recognised by his father as the heir to Ghaznävid

throne, in an episode that seems to reflect his father's own life, Mas'üd fell from favour and was

imprisoned for a short time in Mültan sometime in 411 AH (Nizam, 1993: 400). He subsequently

was able to regain his father's trust, was awarded the governorship of Rayy in 420 AH, and had

succeeded in fully conquering the Buwaihis' Isfahan, preparing for a new campaign against the

Turkmen, when Mahmüd died (Briggs, 1829: 31). Muhammad can only have assumed rule with

the support of elements of the court, and in this sense, his accession reflected the dominant interests

of sections of the Ghaznävid elite at this point, in fulfilment of the perceived wishes of Mahmüd, as

the passage from Muhammad 'Cif suggests. This may have been for a number of reasons, but

clearly the wider interests of dynastic politics came into play; Muhammad was effectively

appointed by Mahmüd's brother-in-law, Amir Ali b. Kuzil Arslän (Briggs, 1829: 31). Muhammad

'Off suggests that the military support for the accession of this prince was lacking (Elliott 1867-77:

vol. 2, section 45), and the events that followed would indicate that Prince Mas'üd was accepted by

the generals and officers as their preferred claimant, as Firishta relates (Briggs, 1829: 31). Fifty

days after Mahmüd's death, a group of nobles and officers, led by one Abü-1 Amir Ayaz (noted as

"son of Isaac" by Firishta and therefore a convert), convinced the household ghuläms to rebel

against Muhammad and seek service with Mas'üd in Nishäpür. This episode is interesting in the

context of this book, as the Prince Muhammad "despatched Sewund Ray, an Indian chief, with a

numerous body of Hindon cavalry, in pursuit of them " (Briggs, 1829: 31). Despite their defeat and

the consequent escape of the rebels to Mas'üd, Muhammad continued to rely upon Hindus for

military support, which would suggest that the household troops that absconded to his brother were

primarily the Turkish ghuläms and that, as suggested previously, the Hindu troops remained loyal to

168


the reigning prince, until his deposition at Herat, some five months into his reign, Shawwäl 421

AH/October 1030 CE (Mohamed Khan, 1835: chap. 5, section 4). Mas'üd's coup was legitimised by

the award of a number of laqabs from the Khalif al-Kädir bi'alläh (Nizam, 1993: 401)

Factional rivalries at court were reflected in the divisions apparent in urban Muslim society between

differing groups that became articulate in local leadership and the contestation of authority with the

dynasty. The process of founding wägfs (Ar. ) or vakifs (Tur. ), had enabled the religious scholars or

`ulamä/ulemä to become increasingly independent of the centralised control of the Ghaznävid state,

as this form of pious endowment provided incomes from monies left by wealthy merchants,

landowners and office-holders. These alliances brought the ulemä into the social and political

sphere as "notables" (aLyän), negotiating with the aýabiyyät or popular factions, the various

schools of religious law, the differing and increasingly popular sufi sects (who provided a more

accessible form of the practise of Islam amongst the populace). The politics of Khorasan saw these

groups competing for leadership with their rulers, and the ability of the former to organise elements

of the local population into militias further enhanced their role (Lapidus, 1966: 347). The

competition that existed between differing urban groups, so palpable during the reigns of Sebüktigin

and Mahmüd and in their manipulation of the heterodox Karämiyya sect in Nishäpür for example,

evolved into increasingly identifying the rapacious demands of the fiscal revenue collectors as a

common threat (Lapidus, 1966: 345).

Part of the impetus for increasing urban organisation was the disorganisation of the rural society,

with whole sections of the rural population resorting to banditry and exactions from the peasantry.

The ever-present ghäzis, so intrinsic to the Indian campaigns, became problematic during the

periods when no major assaults upon the riches of the Hindü kingdoms were being undertaken. The

'ayydrs and hadithän added to the independence of action that these parts of the empire's population

began to exercise, sometimes in alliance with the ulemä but often at their own behest (Lapidus,

1966: 346). An increasingly centripetal tendency emerges during the latter part of Mahmüd's reign

and particularly after Amir Mas'üd assumes power in 421 A11/1 030 CE, the beginnings of a definite

urban identity and culture in counterpoint to the rulers at Ghaznä became apparent (Bosworth,

1963: 254). Mohamed Khan suggests this stark division of the empire's social groups, and the

particular disorder brought by the Saldjüks, when he writes that

" ... in his {Mas'üd's] absence, the power of his formidable enemies had increased; the

property and lives of his subjects were made their sport; and, as one of his nobles said, "though once but ants, they had become adders. " (chap. 5, section 4; Briggs, 1829: vol. 1,

169


section 11)

Despite frequent advice from his ministers and advisers, the Amir continued to pursue a policy of

assaulting the Hindii states, allowing the incursions of the Türknien to plunder the western portions

of the empire and give rein to the aspirations of the previously subjugated Buwaihis, lieutenant-

governors of Rayy, Isfahan and Persian Irak, to independent rule (Mohamed Khan, 1835: chap. 5,

section 4). Mas'üd's obstinacy in the face of his responsibilities, as described by the commentators

on his reign, is regarded as contumacy in the face of divine will, and hence the loss of charisma that

fatally undermines his right to rule. The visitations of pestilence and plague in the Persian lands, in

the year 424 AH/1033 CE was seen in retrospect, as further indication of divine displeasure (Briggs,

1829: vol. 1, section 11). The defeat of Tatiah "an Indian chief' at the hands of the rebellious Ahmad Nialtigin, the governor of Hindustan, prior to the latter's eventual destruction by the Hindu

general Tilak, was another example of the unravelling of Ghaznävid authority amongst its own

competing power-groups. The struggle for control of the entire Hindüstän army, between Ahmad

and the incumbent general Käz Shiraz, underpinned the governor's rebellion using the army of Lahore; in addition Ahmad's success in the field exacerbated the competition between these two and the Amir Mas'iid, and drove him to rely upon Indian commanders and troops to carry out his will. Clearly, the fractured relationship of ruler and elites lay behind the catastrophe at Dandänq. n, at the

close of this turbulent period in 1040 CE.

The incursions throughout Mas'üd's reign had inflicted some shameful defeats upon the Ghaznävid

armies, and local populations were clearly resentful of the inability to defend them against the raids

of the Turkmen (Briggs, 1829: vol. 1, section 11). The situation deteriorated further after 426 AH,

and in 428 AH Mas'üd heard the complaints of his subjects in person when he arrived in Balkh, and determined to eradicate the threat of Tüghril Beg and the Saldjüks in a campaign beginning that

winter (Briggs, 1829: vol. 1, section 11). His plans were thwarted both by the weather and his ignorance of his own commanders' advice. The two major groups of Saldjüks, in Irak and Trans- Oxania required dealing with in this order; those under Jakür Beg had successfully raided Khorasan

and defeated Ghaznävid forces consistently, whilst their "cousins" under Tüghril had constantly raided across the Oxus, only to melt away at the approach of any Ghaznävid expedition, yet Mas'üd "by no means listen to the advice of his officers". Having bridged the river, Mas'üd occupied the territory of the Saldjüks, but to no avail, as they had meanwhile attacked Ghaznä itself, raiding the Amir's own stables and causing much damage to the outskirts of the city, before being repelled (Mohamed Khan, 1835: chap. 5, section 4). In Ramadan 429 AH/June-July 1038, the Saldjüks took

170


Ni*äpür, and Tüghril Beg "sat upon the throne of the realm of Sultan Mahmüd at $ädhyäkh", an

unacceptable affront to Mas'üd, who "when this news reached... [his]... ears, he was beside

himself and extremely anxious and upset. " (Luther, 2001: 37) Despite his earlier intransigence over

the advice from his generals, this was a challenge the Amir could not ignore.

The march to Danddngän led through the almost waterless desert of the Sarakhs region, when the

drought in the area had been exceptional (Bosworth, 1963: 250) Mas'üd, re-occupied the city of

Ni dpür in late 430 AH/1039 CE and surviving the subsequent famine, marched into the Tüs region

to try and secure supplies, but the years of depredations by nomads and finally by exhausted

Ghaznävid armies had reduced the region to bareness; neither bread nor barley could be found and

the Saldjüks had stopped-up the wells, or spoilt them (Luther, 2001: 38). On the passage to Merv,

almost all of the horses died, leaving the heavy cavalry and ghuläms inexpertly on camels, along

with many of the infantry (who usually travelled that way on campaign; see Bosworth, 1963: 113).

In the Saljüq-Näma an interesting story about how Mas'üd wished to stop cagri Beg and Tüghnl

Beg joining forces in TGs, and hastened to the location on a forced night-march. However, he fell

asleep on his elephant and the elephant driver, fearing to awaken him, drove the elephant slowly

and allowed the Saldjüks the advantage (Luther, 2001: 37). The battle itself saw desertion of the

former Turkish household ghuläms of Sultan Mahmüd to the Saldjüks, and left the Hindus as almost

the only troops at Mas'üd's command. Baihaki mentions some 16,000 Turkmen (64,000 including

family members; see Bosworth, 1986: 938), but we have no clear indication of the numbers of

Ghaznävid troops actually present; the march had exacted a terrible toll upon the armies of Mas'üd

(Bosworth, 1963: 250), so it is difficult to postulate how many of the soldiers were likely to be

Indian in origin. The pattern of the Amir's reign would suggest that the majority of those who fought for him, and who consequently suffered the worst casualties, would have been the Hindu

ghuläms, (see plate right of a Ghaznävid Hindu palace ghuläm; Nicolle & McBride, 1982: 27), and

elephant drivers. These troops, with no ties of loyalty other than to the person of the sultan and

without connections of ethnicity or religion that would allow them to ally themselves with the

Saldjük Turkmen, would have been decimated. Arguments have been made that the defeated Hindus

would have been incorporated into the Saldjük armies "Indian troops passed from the Ghaznävid to

the Seljuq [sic] armies" as an opportunistic move, common after the defeat of an enemy (Bosworth,

quoted in Hancock, 2005: 26). Such acquisitions of one's former enemies was, and still is common

practice, and recorded frequently; in this case the shift of allegiances from Mas'üd to Tüghril Beg at Dandängän is a good example. However, the sources do not mention the incorporation of Indian

171


troops into the Saldjük forces, nor is it likely that it took place. The organisation of Saldjük warfare

was clan-based, mobile family units at this point, only later developing what might be described as

the classical Islamic ghuläm structure and then only partially. As Bosworth argues elsewhere, "the

sultans never conceived of themselves as despotic rulers over a monolithic empire, rulers in the

Perso-Islamic tradition of the power-state as it had developed... under the early Ghaznävids. "

(1999) Furthermore, he goes on to write

They had risen to power as the successful military leaders of bands of their fellow Oghuz

tribesmen and at the outset depended solely on these tribal elements. The position of the

Saldjük Sultans was thus fundamentally different from their predecessors in the East... the

Ghzanawids, themselves of slave origin and dependent on a purely professional, salaried

standing army; likewise their opponents in the West, the Biiyids and Fätimids, had come to

depend upon professional, multi-ethnic armies. " (1999)

Unable to sever their connections with the Turkmen tribal elements, even when it became a matter

of dynastic control, the Saldjüks never divested themselves of their steppe origins, nor it would

appear, aspired to do so, maintaining tribal custom in partnership with their role as Muslim rulers

(see Lampton, 1968: 218). No record of Indian commanders or officers comes to light in the Saljüq-

Ndma of Zahir al-Din Nlýdpüri, or comparable sources, in the way they do with al-'Ütbi, Baihaki

and other Ghaznävid narratives. Thus the "Saldjük factor" must be limited to the impact upon the

Indian giqäniyya and ghuläm (see plate left, Nicolle, Hook & Turner, 1998: 38) units at Dandängän

in 1040 CE, and their subsequent role as a "motor force" in driving refugees from this calamity into

Anatolia, as an increasingly composite group fleeing the "savage nation of infidels called Turks... "

(Dostourian, 1993: 44).

172


5.3 "... no sultäns greater or kinder to the reäyä [flock]... than the kings of the house of Saljuq... "

The Persian geographer, Mahmüd al-Kashgari, writing in the 4th century AH/10th century CE,

enumerates the twenty-four tribes of the Oghuz or Ghuzz in his Diwan Lughät al-Turk (Houtsma,

1993: 208; Rice, 1961: 25). Their fascinating, though complex lineage is described comprehensively

elsewhere (see references in the previous section), and it is unnecessary to review here, but it is

important to note that they descended from the Qiniq tribe, a leading clan in the Döküz (Nine) Oguz

and holders of military office (sübaýi) under the great commander of the right-wing of the horde,

the Yabghu (Bosworth, 1999). In the conflicts of steppe society, the Oghuz were one of the groups

that became displaced in the late 3`d century AH/10`" century CE and migrated westwards. The

Saljüq-näma, after praising their treatment of their subjects (quoted above; see Luther, 2001: 3),

records that

... it was learned thus from historians and authors, that Saljuq b. Lugmän was from the

Saljuq family of the Qiniq tribe ... The house of Saljuq was a great family and very

numerous. They had inummerable possessions, complete provisions, and well-organized

regiments and retinue. (Luther, 2001: 29)

The later Saldjüks reconstructed a more illustrious heritage at a later date, as Tüghril Beg's vizier

Abu'I'Alä' Ibn Hassöl (d. 450 AH! 1058 CE) made claims for them to be related to the legendary

Afrasiyäb (Bosworth, 1995: 938). By the beginning of the 5th century AH! 11`h century CE, the

Saldjükids had migrated from their abode in on the Volga steppe, to the Trans-Oxania region, and

had come into contact with the Ghaznävids, to the concern of both the Qarägänid Ilek Khan and the

Sultan Mahüd of Ghaznä (Bosworth, 1995: 938; Luther, 2001: 30; Talbot Rice, 1961: 28). The latter

not adopt the strategy suggested by his Hädjib (a military commander) and arrange for the

amputation of the thumbs of the Saldjük archers (Talbot Rice, 1961: 28). Instead he did require the

depositing of hostages with him, after hearing that the Saldjüks could raise 100,000 troops merely

by sending two arrows amongst the tribes and as many as were necessary if a bow were sent

(Bosworth, 1999; Houtsma, 1993: 208). The conflict between the Ghaznävids and Saldjükids

became exacerbated when Isrd'il (or Arslän), the leader of the Saldjüks at that time, was taken back

to Ghaznä as Mahmüd's guest and thence imprisoned in the Kälandjär fortress in MUltan, sometime in 1025 CE. This event is recorded in the Saljüq-n . ma as an act of supreme betrayal by Mahmüd,

trusted as he had been by Arslän Isra'il, his son Abu '1-Fawäris Qutlumu§ and "... three hundred

youths of good appearance and pleasant form... " (Luther, 2001: 30).

173


These measures failed to curb the turbulent behaviour of the rest of the clan, or what Bosworth

describes as the "... unimportant chieftains of what must have been rather shaggy bands of Turkish

nomads. " (2001: 8) Despite their relative importance in the Islamic world at this point, their

withdrawal from Ghaznävid jurisdiction into Khoräsän, Irak and Adharbaidjän caused extensive

disruption to Muslims elsewhere, as the Saldjüks began to raid extensively (see al-Baihaki,

1862: 544). Matthew of Eddesa records a raid in the year 467 of the Armenian calendar (1018-1019

CE), by Cagri Beg, as the first appearance of the Turks, causing widespread fear and distress

(Dostorian, 1993: 47). This raiding remained a pattern throughout the rest of Mahmüd's reign, for

those called the `Iräki Ghuzz by al-Baihaki, in contradistinction to those known as the Trans-Oxania

Ghuzz. These latter, referred to as the people of Tughnl Beg Muhammad b. Mikd'il b. Saljuq by al-

Kashgari, remained in Nür Bukhdrä, under the former's leadership. As pastures and resources grew

scarcer in this region they were expelled from Khwarizm for their depredations, and the Saldjükiyän

and Inäliyan sought permission to move into Farona and Nasä as mawäli, which was denied them

by the Amir Mas'üd (1035 CE, see Houtsma, 1993: 209). Despite the offer of sending more

hostages to the Ghaznävid court, the Amir was intransigent in his rejection, and the Saldjüks began

raiding Khoräsän in earnest, also defeating a punitive expedition against them in 426 AH/1035 CE

(Bosworth, 1963: 242). Eventually forced to recognise what was by then a de facto control of the

region around Merv by the Saldjük clan and their Turkmen tribes, Mas'üd only maintained any

room for negotiation by exploiting their internal divisions.

The deterioration of relations, together with the perfidious behaviour of Mahmüd and the perceived

arrogance of his son, are cited as justification for the demise of the Ghaznävids, their loss of

authority and rulership through charisma (see Bosworth, 2001: 8; Luther, 2001: 35). The notion of

charismatic rule, the particular "good fortune" attached to a dynasty, is of course, well-known from

Central Asian nomadic societies, and most especially the Mongols (Hartog, 1989: 35), and it is

interesting to see Zahir al-Din Nishäpür deploying it in the pursuit of legitimising the replacement

of one Sunni Turkic dynasty by another (Luther, 2001). The result of this increasing acrimony was

eventually open warfare between the Saldjtiks and their Ghaznävid overlords, in which the former

were consistently superior, "... their lightly-armed and highly-mobile cavalry force proving more

than a match for the more heavily-armed but cumbersome Ghaznävid army... " (Bosworth, 1995:

938). Clearly at his point, the Saldjükid clan and their Turkmen followers became extremely

influential in the Muslim world, as they were able to inflict one of the most serious defeats upon a

major Islamic polity, thereby altering irrevocably the development of the mediaeval Arab and

174


Christian lands of the whole region. More importantly for this book, the question of the so-called

"Saldjük factor" and the suggested role of the Turkmen in the early history of the Romani peoples,

becomes clearer, in relation to the events of 431 AH] 1030 CE.

The primary contention I am making here is that the Hindu regiments of the Ghaznävid army would

not have even been in Trans-Oxiana, on the plain of Dandanqan near Marv, had it not been that they

were a major and important part of the Ghaznävid army and palace guard. The continuing challenge

of the Saldjük Turks had been less of a priority for the Amir to this point, than had the continuing

campaigns into India. The reasons for this are not made clear in the narrative sources; indeed

Baihak and others clearly ascribe to fate the choices made by Mas'üd; "The star of the King's

fortune had now reached its zenith, and he would by no means listen to the advice of his officers... "

(Briggs, 1829: 97). Yet there are other intimations about the reasons why the Saldjüks failed to

attract the attention of their overlords. As nomadic, uncultured tribes people living on the fringes of

the Ghaznävid Empire, their seemingly insincere requests for settlement in the borders of the

empire as mawlä, clients of the Amir, whilst ravaging the settled lands periodically, were regarded

as indicative of their inferior status. In 425 AH/ 1034 CE, the generals Hussein and Bogtüdi were

sent against the Saldjüks, at that time in the Niýäptu region. A petition from the Turkmen to the

generals stressed that the Saldjüks and their followers were "the King's servants" who desired only

to live at home with a subsidy from the royal coffers, and be led as ghazis in warfare for plunder.

They wished no harm on his subjects, only his enemies (Briggs, 1829: 95), but the message was

clear enough, it was the traditional strategy of Central Asian nomad confederacies in extracting

subsidies by threat (Barfield, 1989: 24). The generals' reply was "imperious and haughty",

informing the Saldjüks of their duty to submit to the authority of the Amir, cease their depredations

or suffer the penalties (Briggs, 1829: 95). This rejection meant that the Saldjüks attacked and would

seem to have adopted another classic nomad strategy, the feigned retreat, to inflict a defeat upon

their imperious opponents, Hussein being captured and Bogtüdi fleeing with news of the disaster to

Ghaznd.

Here is the key to the relationship between the Saldjükids, their Türkuren followers and the Ghaznävids. The descendants of the Oghuz are recorded as operating in ways that are clearly derived from Central Asian nomadic practice, in relation to a more sophisticated polity, where the

complexities of their dependence upon a strong empire were as necessary as their stable social

structure (Barfield, 1989: 20). That the Saldjiiks developed the latter is obvious in that they had the

ability to build both their own imperial system, and maintain the traditions and customs of the

175


steppe when these did not conflict with political necessity (Bosworth, 1995: 939). Without that

stability, this would not have been possible, and it was clearly not apparent in their capture of

Ni*dpür in 429 AHl1038 CE, which they could only hold temporarily. This suggests that the

experience of this event, and the defeat of the Ghazndvids in 431 AH/1040 CE forced rapid changes

in the social structure of the Saldjüks, encouraged by the acquisition of significant experience from

those renegades who abandoned the Ghazänvid amir. As with the experience on what Barfield calls

"the perilous frontier" of northern China, the administrative collapse of the sophisticated polity (in

this case, western Ghaznä and Khoräsän), forced the Turkmen nomads to replace the previous

system with their own revitalised and hybrid organisation (1989: 36). The organising principle of

the Turkmen was the logic of charismatic rule, and the particular "morning of fortune" of the

Saldjükid clan, as Zanir al-Din Niýdpüri's Saljüq-Näma tells us, it "... shone from the pages which

told of their circumstances, and the splendour of good fortune glinted from the leaves which related

their condition" (Luther, 2001: 35). A more prosaic explanation (though admittedly less poetic), lies

in the dynamics of Central Asian steppe-sown politics.

The implications of this for Romani history are profound, in that had Mas'ud thought less of

imposing his direct will upon the Oghuz Turkmen, rather than adopting the strategy that the

Saldjüks themselves developed with regard to the turbulent tribes people, of utilising them in

raiding external lands, he may not have lost the western empire. The Saldjüks themselves may have

been effectively deployed as governors of the border regions, thus continuing their dependence

upon the dynasty, and forcing them to address the depredations carried out by their erstwhile

cousins. In effect, the defeat at Dandängän in 431 AH brought large numbers of Hindus into the

region as troops and servitors whose dependence upon their Ghazndvid patrons was severed by the

latter's abandonment of the survivors, and withdrawal to India. It would not be overstating the case

to say that without the Saldjük defeat of the Ghaznävids in 1040 CE, there never would have been

any Gypsies, as these are the descendants of those Hindu warriors, fletchers, armourers, grooms,

blacksmiths, tent-makers, cooks and sundry others attached to the Ghazndvid army that had lived

through the campaign and marches.

That these survivors may have included few warriors is highly likely; traditional Rädjpüt tactics

meant that the Hindü ghuläms would have relinquished their lances, dismounted and, binding their

shirt-tails together, fought to the last with maces, axes, daggers and clubs. The fiercest of the

combat would have been around Mas'üd's elephant, where the Indian troops would have

congregated to defend their commander, the death or disappearance of the Amir being taken as a

176


sign by both sides of the defeat of the Ghaznävids (Rizvi & Burton-Page, 1986: 202). NNýdpün tells

us that the weight of Mas'üd, together with his weapons, and armour were too great for a horse to

bear, but that he was pursued by several Turkmen and forced to mount a horse to defend himself

alone, presumably after his elephant had died or he was struck from it, and his guard were all dead

(Luther, 2001: 38). Despite Mas'üd's "Rustam like blows", the battle was lost, as was the western

empire, but this passage indicates that the battle would have been deemed lost once Mas'tid was no

longer mounted on his elephant, and that the likely carnage amongst the Hindu ghuläxns was

appalling.

The "Saldjük factor" is crucial then, but not in the way that has been identified by other scholars;

there were no large-scale incorporation of Hindu warriors into the Saldjükid forces after Dandängän

(Bosworth, quoted in Hancock, 2005: 27), and indeed, the mechanism by which such personnel

would then become independent groups of "Egyptians" arriving in eleventh-century Constantinople

is not explained in this notion (Hancock, 2005: 25-27). As captives and slaves, the Indians may

have ended up in the markets of the region, being sold as domestics and labourers to individual

owners, but again such a scenario fails to address the problems of how these individuals would have

made their way to Byzantium (presumably as escapees), and reformed as distinct "units" that

arrived in the capital sometime later, through chaotic and dangerous territory swarming with

Turkmen bands and refugees from their attacks (Dostourian, 1993: 144). Finally, the linguistic

evidence fails to support such a contention, as there are frequent references in the literature to the

absence of Turkish loan-words in the Romani lexicon (see Fraser, 1992: chap. 2). The use of Persian

amongst the Saldjük elite was something that happened after the conquest of Khorasan, as was the

development of a ghuläm army (as suggested above). The language of command developed into

Perso-Turkic, or "Old Anatolian" (as did the language of administration), but the renegades were

from the Turkish ghuläms associated with Sultan Mahmüd, not the Hindu ghuläms and generals that

Amir Mas'üd relied upon to replace the influence of his father's commanders (as testified by

Mas'üd's refusal to listen to their advice; Briggs, 1829: 96). The common tongue between these

renegades and their new overlords was, of course, Turkish, not Persian, and the language of Islam,

Arabic. Both these would have become incorporated in some terms into the Romani lexicon, had

the Hindus been a subsequent part of the Saldjükid forces. In the conquest of Armenian Anatolia,

the Hindus would have been associated with the "blood thirsty beasts" that Matthew of Edessa

records, and extremely unlikely to have developed a close enough association with the Armenian

survivors of these raids, to account for the influence of this language upon the basic Romani lexicon

177


(Dostourian, 1993: 47).

The Saldjüks, or more correctly their shamanistic, tribal, nomadic Turkmen followers that were the

main military component in the early, Saldjukid polity (see Bosworth, 1999) were the motor of

change, the literal driving force behind the proto-Romani migration. These tribes people almost

"herded" the survivors, pushing ahead of them the remnants of the defeated Hindu giqäniyya

regiments after the bloody conflict of Dandängän in Ramadan 431 AH/May 1040 CE. Their path

would have followed the "silk route" from Merv, Nshapur, the southern Caspian coast, Rayy and

into the Ädharbaidjan and Arminiyan lands, the ancient roads of East-West communication and

trade. In the chaos of the borderlands between the emerging Saldjükid empire and the contested

lands of Arminiya, the groups of Hindus, Persians and others moving on the roads of the Iranian

lands passed unrecorded, and it is only possible to suggest their passage through informed

speculation. The Turkmen flooded into Khorasan, Arminiya, and eastern Anatolia as a result of the

complete collapse of the Ghazniavids in their western empire, ahead of them an increasingly

composite group of Indians, Persians, Armenians, Greeks and other "refugees" fleeing from what

Matthew of Eddessa describes as the "crazed and pernicious nation of the Turks" (Dostourian,

1993: 101). The practice of "chasing" defeated populations ahead of advancing forces in Mongol-

Turkic warfare, was designed to instil terror and disruption in the threatened communities. The use

of the hunt or §ayd, as a method of training for war was common to most Mongol-Turkic groups, including the Ottomans, who carried out large hunts as "training" during the periods between

campaigns, and this form of warfare, or maniere de combattre appears in later accounts of nomad

tactics (Vryonis, 1971).

The Saldjuks were not interested per se in Anatolia initially, except to distract the wild and woolly Turkmen from ravaging the Persian lands and causing enormous distress for the sedentary

population (which they often did anyway), by sending them into the lands of Rüm. The control of the nomads was continually problematic, and ultimately undermined the fabric of the Great Saldjuk

Empire. Frequently insubordinate, the Turkmen tribes people were often unreliable in warfare,

either leaving campaigns when they had gathered "enough" booty, or fleeing battlefields altogether (Bosworth, 1986b: 197). Their campaigns against the Armenians and Byzantines were in the nature

of deep, and highly destructive raids in search of plunder, slaves and as exemplary expeditions to

convince their own vassals of the power and might of the Saldjük sultäns (Cahen, 1968: 72). The

view from the Armenian side was, of course, very different, and the chronicles of Matthew of Edessa and Michael the Syrian, amongst others, record the apocalyptic terror, the perception that the

178


end of the world had come. When King Senek'erim sought an explanation for the arrival of the

"infidels" in 10 18-19 CE, he found amongst the prophecies of the Armenian vardapets confirmation

of his despair, in the words "... at that time they will flee from the east to the west, from the north to

the south, and they will not find rest upon the earth, for the plains and mountains will be covered

with blood. " (Dostourian, 1993: 45)

It is Acton's suggestion that it is an event like this that is referred to by Romanies in mediaeval Italy,

when questioned about their origins (Acton, 2004: pc), suggesting the conquest of the Armenian

kingdom of Cilicia, or western Armenia by the Ottomans in the final years of the fourteenth-century

and early fifteenth-century. In examining the context of this period, it is my contention that it is the

destruction of the centres of the mediaeval kingdom of Arminiya, Akhalk'alak and Ani, during the

years 513 AE (Armenian Era)/1064-5 CE, that these Romanies referred to. The calamitous events

would prove the second major defeat of the ancestors of the Gypsies by the Saldjüks, and all but

eradicate the remainder of the original Hindü warrior caste, whilst leaving their descendants with a

"folk-memory" of this apocalyptic scenario. It is to this that we now must turn.

179


5.4 "... those wicked and abominable children of Ham... " the irruption of the Saldjiiks into Armenia

The first of the Turkmen groups to move westwards to begin raiding the Armenian lands, as

Matthew of Edessa records, did so in 467 AE/1018-19 CE (Dostourian, 1993: 44), although this is

contested by more modem historians (Bosworth, 1995: 938). Certainly, there is no reference in the

chronicle of Vardapet Aristakes Lastivertc'i, to this event. The Armenian cleric Matthew describes

the attack in the most calamitous of terms as the first appearance of "... the bloodthirsty beasts... the

savage nation of infidels called Turks... " (Dostourian, 1993: 44). The timbre of the sources is

however similar, in that this period is one of dreadful events, of appalling disasters that befall the

Armenian people at the hands of "foreign people around us" as Aristakes suggests in the title of his

narrative (Bedrosian, 1978). The rapacity of the Romans (Greeks) of Byzantium is matched only by

the cruelty of the Persians (Turks), in these sources. In the same year that the Turkmen first despoil

the territories around Lake Van, the northern parts of Armenia and southern Georgia was under

assault from Basil, Emperor of Byzantium (976 - 1025 CE), who proceeded to "make the well-

cultivated country devoid of people, a devastated wasteland... ". The king of that country, Georgi

had refused to submit to the Emperor, and suffered the consequences; "... Alas this narration, alas,

this wicked deed" (Bedrosian, 1978: 14). The destruction of the Armenian kingdoms is related as a

result of the perfidy of the Greeks, and the wrath of the Saldjüks, as a sign from God that the

behaviour of the Armeniansas Christiansis at fault (Bedrosian, 1978: 63), and in this sense the

sources have to be examined with this in mind, produced as they are by Miaphysite clerics

embattled by Orthodoxy and Islam.

The resistance to the first, of the Turkmen incursions was led by the son of King Senek'erim, David.

They were fierce, though ultimately unsuccessful, and resulted in many casualties due to the arms

and appearance of the unknown Saldjiik forces, "... until that time the Armenians had never seen Turkish cavalry forces. When they encountered these Turks, armed with bows and having flowing

hair like women, they found them strange looking. " The Armenians were forced to retreat to Lake

Ostan (Van), despite inflicting great damage with their swords (Dostourian, 1993: chap. 48,45)

Matthew laments that the wrath of God had been awakened, "the fatal dragon with deadly fire rose

up and struck those faithful to the Holy Trinity. " (1993: chap. 47,44). The Saldjüks, led by Ibrahim

and Kutulmiý, mounted a series of raids that extended their range deep into Armenian and Georgian

territory (Talbot Rice, 1961: 30-31), causing widespread slaughter in the various town they

assaulted. Matthew records that the Emperor Constantine Monomachus, in the years 497-498

AE/1049-50 CE, "perfidiously and by false oaths" forced the ceding of the Bagratid kingdom,

180


removed the Armenian troops (in order to defeat the rebel Leo Tornices; see Vryonis, 1971: 86), and

left the country unprotected. The Byzantine commanders, referred to as "eunuchs" by Matthew, did

nothing to stop the forces of the "infidels" falling upon the populous and wealthy, but unfortified

city of Artsn (Artze) near Erzerum, which resulted in huge slaughter of the townspeople and

garrison (Dostourian, 1993: 76). The Byzantine governor of the region had in fact to resort to a ruse

to defeat the Turkmen forces, falling upon them whilst they pillaged the Byzantine encampment at

Kapetru, as they were not sufficient in strength to meet the Turkish forces in open battle. Whilst

awaiting reinforcements from Georgian vassals, the Saldjüks were able to lead the sack of Artze,

with the following confrontation being indecisive (Vryonis, 1971: 86). Both Aristakes and Matthew

refer to this event as the beginnings of a major assault on the commercial centres of eastern Asia

Minor:

"In the same year, the gate of Heaven's wrath opened upon our land. Numerous troops

moved forth from T'urk'astan, Their horses were as fleet as eagles, with hooves as solid as

rock. Well girded, their bows were taut, their arrows sharp, and the laces of their shoes were

never untied [i. e., they were always on the move]. Having arrived in the district of

Vaspurakan, they pounced upon the Christians as insatiably hungry wolves devour their

food. Coming as far as the Basen district and as far as the great estate called Vagharshawan

they demolished and polluted twenty-four districts with sword, fire, and captive-taking. "

(Bedrosian, 1978: 66)

Matthew writes that "... going forth full of rage... like a river swelling up with tempestuous rage

and like a beast crazed by its bloodthirsty nature... " the Saldjilks subjected Armenia and "Albania"

(Abkhazia) and "Iberia" (Georgia) to the sword and enslavement (Dostourian, 1993: 101). Vryonis

also identifies this as the beginnings of the destruction of "the important urban centres of Byzantine

Anatolia", occasioned by the inept military policies of Constantine IX Monomachus (1042-55), that

resulted in the dissolution of indigenous troop units in these themes (1971: 86).

The picture of this period is one of increasingly desperate circumstances in eastern Anatolia, where

defences were being irretrievably weakened by internal Byzantine disorder and rebellion, conflict

with the Armenian and Georgian kingdoms of the region, and raids from both the Saldjüks, their

Turkmen followers, and a host of other nomads into the Thracian theme (Ostrogorsky, 1986: 334).

In order to defend the Balkan frontiers against the Pe6enegs or Patzinaks, it was Constantine's

policy to withdraw troops from the Macedonian and eastern Anatolian themes, as a consequence of

the disloyalty of the western commanders (Vryonis, 1971: 87; Ostrogosky, 1986: 334-5). Michael

181


Attaliates, a Byzantine statesman and historian, describes the situation as happening

... because of the venality of the king. It was a large army in Iberia that was sustained from

the taxes of those countries [Armenia, Georgia and Abkhazia], the king seized the taxes, and

by this means he lost a great deal of power, because not only did he forsake his allies, but

transformed them into powerful enemies, adding them to the power of his opponents

[Saldjüks], so making them completely irresistible... (Koukounas, 2004)

The point here is that the eastern Byzantine Empire was in chaos, and that soldiers were in short

supply for the defences. Given the drafting of all and sundry into the Byzantine armies at the time,

had any contingents of Hindü warriors been available, it is highly likely they would have been

employed by the Emperors, yet none appear in the sources. The Byzantines used a multi-ethnic

army at this stage, even paying the Peeeneg mercenaries, however unreliable they proved as frontier

garrisons (Ostrogorsky, 1986: 334). Able to secure some sort of alliance through the award of court

titles and land grants, the Peeenegs were only one of a number of peoples "bought" by the imperial

treasury in an effort to make up the military levy, heavily reduced by battles between rebels and the

emperors, which were to grow worse with the rebellion of Isaac Comnenus and the remaining

Anatolian aristocracy in the years 1056-57 CE (Vryonis, 1971: 87). The suggestion from some

Romani Studies scholars that Indians were part of the military machinery of Byzantium, deduced by

their appearance in much later Venetian sources as companies defending the Serene Republic's

island territories against the Ottomans, cannot be borne out (Lee, 2004: pc).

The alternative notion of the remaining Indians, Persians and others (or what we may now describe

as the proto-Romani), being involved in the defence of Armenia may have more substance to

support it, though it is still extremely speculative. The end of the Armenian Bagratid Kingdoms

came in the year 513 AE/1064-65 CE. Alp Arslän Muhammad b. Cagri Beg (455 - 465 AH/1063 -

1073 CE), whom Matthew of Edessa mistakenly records as brother of the Sultan Tüghril Beg,

collected a vast number of troops from Persians, Turks, "all of Khüzistän [south-western Persia] and

Sijistän [eastern Persia, bordering upon India]". Marching to Armenia and "Albania", he subjected

the whole region to fire and sword, enslaving many and causing countless deaths "at the hands of

the crazed and pernicious nation of the Turks" (Dostourian, 1993: 102). His major objective was the

most important Byzantine-Armenian city in eastern Anatolia, and "the Sultan came like a

threatening black cloud, and descending upon the royal city of Ani, surrounded it completely on all

sides like a vicious serpent" (Dostourian, 1993: 101), forcing the population into the confines of the

walls with his first violent assault. Perceiving that the entire population were within, the Saldjük

182


sultan ordered a prolonged investment of the city, whilst the inhabitants began to fast and pray. The

city held for some time, except for an area described by Matthew as being on one side, about a bow-

shot away and of temporary dwellings, which was destroyed quickly by Turkish catapult-fire

(Dostourian, 1993: 103). The tragedy unfolded when the garrison of "Romans" (Greeks) withdrew

into the citadel, presumably leaving the walls to the citizens, causing mass panic. The Saldjüks

seem to have lifted the siege just prior to this, but realised the city was now vulnerable and renewed

their assault, "cutting down great numbers of them [the inhabitants] like green grass and piling up

their bodies... " (Dostourian, 1993: 103). In a curious reference Matthew tells us that the cathedral

was desecrated, the cross torn down by a Turk who climbed the dome and also

"... hurled down the crystal lamp onto the floor of the cathedral, causing it to be

extinguished; Smbat the Conqueror [King of Ani, 977-989] had this lamp brought from

India, along with an indeterminate quantity of precious objects. When the Sultän learned the

fate of the lamp, which was crystal and without equal, he was distressed. " (Dostourian,

1993: 104)

The population that survived this apocalypse were enslaved, brought chained before Alp Arslän,

"beautiful and respectable ladies... innumerable and countless boys with bright faces and pretty

girls" were carried off to the slave-markets of Persia "... together with their mothers", whilst priests

were burned alive or flayed (Dostourian, 1993: 103). These types of narratives are common enough

from accounts of Viking raids on the holy isle of Lindisfarne in the ninth century (see Sawyer, 1994:

78-9), to the accounts of Mongol attacks in the thirteenth century (Marshall, 1993), and illustrate

the psychological impact of these events upon communities, interpreted through the prism of

religious belief. Nevertheless, the "folk memory" of events, captured in the later chronicles of

Matthew and others, finds its counterpart in the accounts of Romanies in Europe in the fifteenth

century. I would suggest that the fall of Arni and Artsn in the middle of the eleventh century, are the

substance behind these "tales" (see Fraser, 1992: 87; 1990: 51).

The possibility that relatives of the Hindus who fought at Dandangän (1040 CE), also were caught

up in the destruction of Armenia twenty-years later, may lie in the description of the dwellings

outside the Ani city walls (see above). The chaos of eastern Anatolia was not just limited to that

related to Byzantine and Saldjük devastation, as the petty principalities of the Bagratid kingdoms

were also involved in conflicts with each other. The collapse of central authority after the initial

abdication of the King Senek'erim in 1019 CE, the competition between various petty princes

accelerated, exacerbated by the interference of the Byzantine Greeks (see Bedrosian, 1978: 61-68;

183


108-110, for examples). One of the outcomes of this situation was the increasing use of mercenaries

by all those involved, and these were frequently Turkmen, Persians, Dailamis, Franks, Normans,

Varangians and, I would suggest, the remnants of the ghuläms and other Hindu Ghaznävid troops

after 1040 CE. The description in Matthew of the community outside the walls of Ani in 1064 CE,

bears a similarity to that of pilgrims encountering "Egyptians" encamped outside Modon in the

mid-fourteenth century (Fraser, 1992: 45), in temporary dwellings beyond the city walls. The

prospect that these were the habitations of proto-Romani groups must be considered carefully

however, as there are no direct references that would allow us to positively identify this group at

Ani in 1064 CE.

There is one further intriguing reference in Matthew of Edessa's Chronicle that may indicate the

presence of the remnants of the Indians from Ghaznä. The curious story of a caravan is told in

Matthew's chronicle (Dostourian, 1993: 148-149), arriving in Antioch "twenty years before this

time [1064 CE]" when this group of people described as "from the East" set up in the market place

of the city and began to "make merry". The Antiochenes pounced upon them and beat them,

ejecting them from the city; we are not told why. The men of the caravan, eighty in number,

retaliated by fighting with truncheons and forcing the Antiochenes from the city gate at Sewotoy to

the Church of St Peter, where the townspeople swore on the Gospels to leave the caravan in peace,

and "the caravan returned to its place of origin" (Dostourian, 1993: 149). The caravaneers were

trading in a variety of goods including died fish, but not apparently especially valuable

commodities, so it is unusual that the Antiochenes set upon them, as it would not appear to be for

economic motives. It isn't possible to do more than indicate the curiosity of the this story, set as it is

after the description of the destruction of Ani, and the strangeness of the circumstances that would influence Matthew to relate it.

In summary, the evidence for the presence of proto-Romani groups in the Armenian lands is

circumstantial if it could be claimed that there is any indication of them at all. The logic of the

narrative I have constructed previously would suggest that it is highly likely that there would have

been a passage through these lands at this point in their history, and the likelihood of this group or these groups being involved in the events described in the sources is also high. It cannot be argued

with any certainty however, and must remain conjectural, even if informed by the context of the

wider history of Romani people at this time. It may be that further research being undertaken in the

archives of the Armenian Patriarchate in Jerusalem at present, could further inform these perhaps, unwarranted speculations, but the generalised nature of the sources that essentially portray these

184


dramatic events as a battle between good and evil, may not allow us to do more than make

suggestions as to the possible experiences of those people who emerged as "Egyptians" in the late

eleventh century in Constantinople.

185


6.1 "Accidents, evil omens, and ugly spectacles... " the collapse of Byzantine Anatolia 1071 CE - 1176

The establishment of the Seljuks of Ram (as a cadet branch of the Great Seljuk Empire) in Anatolia,

was a direct consequence of the defeat of the Ghaznävids at Dandanqan (431 AH/1040 CE). The

impact of this for the Armenians at Artsn in the year 498 of the Armenian era (1049-1050 CE),

again in 513 AE/1064-1065 CE at Ani, was briefly examined in the previous section. More

importantly, the defeat of the Byzantines at Manzikert (1071 CE) and again at Myriokephalon (1176

CE), profoundly influenced the development of the region (see Langdon, 1992, Bosworth 1991,

Savvides 1981 and Vryonis 1975 & 1971). In particular, these developments are most apparent in

the significant change in the ethnic balance of Anatolia in the wake of catastrophic defeats, military

and political collapse, and the irruption of thousands of Turkmen tribes-people into the region of

eastern Anatolia in the period. The sources may have stressed the extent of destruction and

disruption caused by these nomadic peoples, as they were frequently compiled by ecclesiastical

authors such as Matthew of Edessa (Dostourian, 1993), Bar Hebraeus (Lane, 1999) and Aristakes

(Bedrosian, 1978), who saw the motivation for these raids for booty and slaves as divine wrath and

punishment, but the picture of change during the establishment of Turkish beyliks (frontier

emirates) must be evaluated carefully.

The notion of the confrontation between these new invaders (the Saldjüks and their Turkmen

followers) and the Orthodox and Oriental Christian populations of Asia Minor/Anatolia and the

Levant as one of incorporation (as happened later under the Ottomans) must be revised. As Vryonis

(1971: 155-161), and others have noted, the antagonism towards the Oriental Christians from the

Orthodox Greeks of Byzantium was such that many Georgian and Armenian princes and aristocrats joined the Saldjüks in their campaigns against the Rhomaioi, as the Greeks were called (after

Romans), utilising the disruption caused by Byzantine defeats to establish independent

principalities (Canard, 1986: 638). The complexity of relations affected the changes taking place in

the region in many ways, beyond the simple dichotomy of Islam and Christianity in conflict (Vryonis, 1971: 92-3). The predominance of the Greek population of Asia Minor (Anatolia),

amongst the complex mix of Armenians, Syrians, Hebrews, Indians, Arabs and others was

undoubtedly altered by this irruption, and henceforward, the Hellenic nature of Anatolia would be

steadily undermined, although the process took place over many centuries and only ended with the

population transfers of the 1920's. The particular circumstances of eastern Anatolia or Byzantine

Asia Minor, were apparently far more dramatic however, with wholesale depopulation, destruction

186


of cities, towns and villages, decimation of livestock and the eradication of a centuries-old Oriental

Christian culture.

In this context however, the likely adaptation of the Romani language to the environment of

Anatolia during this period would appear to support the more general picture. Historians who

challenge the suggestion that Hellenic culture was eradicated and the Greek population over-run by

Turks, a view especially popular amongst more nationalist Turkish scholars, cite the continuance of

Orthodox communities as organised confessionally and administratively. The patriarchates of

Constantinople and Antioch show a predominantly Orthodox population were still in place until the

thirteenth century (Korobeinikov, 2003: 197-214). The number of loan-words from Greek

concerning counting and metal-working, amongst other subjects, is very high, whilst it has

influenced grammar at "fundamental" levels (Hancock, 2005: 29). This presence of Greek in the

basic Romani lexicon, strongly suggests interaction with a continuing urban Greek population with

whom proto-Romani people needed to trade and exchange, either over a period long enough, or

important enough to permanently establish these elements in the language as it evolved. Hancock

suggests that in the multi-ethnic and linguistic milieu of Byzantium, children born into what he

describes as the Rajputic speech-community, would have been exposed to variety of languages all

around them (2005: 28). Here it is my intention to suggest the kind of communities that might be

described under the label, proto-Romani, Rajputic-speaking people.

I argue specifically that the proto-Romani, Rajputic-speaking groups emerged as three related

identities in the period between the destruction of Ani (1064 CE), and the second major defeat

inflicted by the Saldjüks upon the Byzantines, Myriokephalon (1176 CE). These identities were

`forged' from elements of groups sharing the Ghaznävid koine, that might be described as early

proto-Romani, Hindü and Khoräsäni displaced by the defeat of the Ghaznävids at Dandängän, near

Merv in 1040 CE (now Marv in modern Turkmenistan), and later, the defeated Armenians after the

loss of Artsn (1049 CE), Ani and Kars (to the Byzantines) in 1064 CE. The resultant group was no longer an organized military force in the way Indians giqaniyyä units were in the Muslim armies of

the Ummayads or Ghazndvids, as the elite ghuläm warrior components had been decimated by

defeat, leaving the survivors from assorted sword and shield-bearers, grooms, armourers, smiths, bakers, cooks, tent-makers, wives, children and entertainers and others that made up the mediaeval

army. A part of these proto-Romani emerged as Rhomiti or Romitoi, as these people were later to describe themselves in Epiros, to those curious of their origins (Fraser, 1992: 53-54; Frescobaldi,

1818: 72). The exact composition of whom emerged as Romitoi groups in eleventh-century CE

187


Anatolia is of course, impossible to trace, but given the picture that emerges from the evidence of

blood-groups and other genetic evidence (Kalaydjieva et al 2004; Iovitä and Schurr, 2004; Potts,

1976; Sunderland, 1976), would clearly suggests an intermingling of various peoples from outside

these core groups as they moved through Persian territory westwards from Merv. I suggest that the

Romitoi came to share a number of characteristics, namely a clan-based social structure,

commercial nomadism as an economic strategy based upon the available skills amongst the group,

and elements of culture derived from the Indian "core" membership. This hybridised set of practices

and beliefs also acquired elements from the environments through which these groups moved,

especially Khoräsän, Adharbäidjän and Arminiya, connecting Hindu symbols and associations, with

Zoroastrian fire-worship, Georgian and Armenian Christianity and Turkmen shamanism. In

acquiring the Ghaznävid koine that would develop into Rajputic, with membership of the group,

Persians, Armenians and others added to the composition of both the proto-Romanies and their

evolving lexicon as they moved into the western lands. At this point, I would suggest that the

uncertainty of existence and disruption to trade, communication and transport in the region meant

whatever goods and services were provided in the villages and urban centres of the region that they

passed through, these were small-scale, rudimentary and unsophisticated. Only with increasing

prosperity and security in the period of Saldjük-Byzantine relations in the thirteenth century, would

the kinds of goods and services develop into more sophisticated and diverse types. This group, or

groups represents one element that survives the atomisation of mediaeval Armenia.

Another group can be seen to have emerged in the dissolution and destruction of the mediaeval Bagratid kingdoms in this period. As Matthew and others record, the roads and countryside of Anatolia was filled with refugees and runaways from the primary targets of Turkmen and Saldjiik

ravages. The aftermath of Manzikert (1071 CE) saw desolation and devastation across the whole of Asia Minor (Anatolia), and "the Oriental peoples... [Armenian, Georgian and Syrian Christians]

... began to decline, and the country of the Romans... [Byzantium]

... became desolate"

(Dostourian, 1993: 143). This state of affairs resulted in "very important and illustrious personages

- nobles, princes and stately ladies - roamed about begging... Because of the famine and vagabond life, there was great mortality throughout the whole land... " (1993: 144). Here lies the solution to

another of the puzzles in Romani history; the appearance of bands of "Egyptians" in early fifteenth

century western Europe, led by "dukes" and "counts". The terms doux and komes, were two

common military titles awarded frequently to Armenians in later Byzantium (Canard, 1986: 638),

the former meaning military commander in the period 10th-12th century CE, the latter being given

188


to subaltern officers at this time (see Kazdhan, 1991: 659,485). The term doukes is also, of course,

associated with the lineage of Doukas of possibly eastern, Armenian origins according to Psellos

(Sewter, 1952: 254-55; although Kazdhan argues this is impossible to prove, 1991: 659), that came

to rule Byzantium under Constantine X Doukas (1059-1067 CE) and Michael VII Doukas (1071-

1078 CE). The vagabond "princes, nobles and stately ladies" were, I suggest, the ancestors of the

later leaders of the bands of Egyptians, by this time ethnically indistinct from their followers, but

maintaining an elite status to differentiate them, as we see described in the Hildesheim and Parisian

records (Fraser, 53-54,66-67). Furthermore, it is not inconceivable that "Duke Andrew", leader of

the Egyptians was in fact an Andronikos Doukas, a descendant of those Doukai who had been

nobles in Armenia and emperors in Constantinople, maintaining a hereditary linkage and title.

Such a suggestion leads me to make two other primary contentions here; the 15t1 century Egyptians

recorded in western Europe, were descended from those survivors of the Byzantine dismantling and

Saldjük destruction, of the Bagratid kingdoms. As previous governors with their retinues they were

able to establish themselves in Lesser Armenia (Cilicia and the Taurus mountains), and later were

incorporated into the Fätamid Khalif'a of Egypt, where they were regarded with some favour

(Canard, 1986: 638). The second contention here is that the atomisation of the Armenian kingdoms

led to the shattering of the Hindü-Khoräsänian, proto-Romani group, with leaderless remnants

fleeing to the west and Constantinople, whilst others were cut off by the Turkmen incursions and

migrated into Karabagh, Georgia and the Caucasus. The latter would develop a version of Romani

heavily inflected with Armenian, called Lomavren, whilst the former were to destined to become

marginalised, itinerant fortune-tellers, bear-leaders and sooth-sayers whom we find featured in the

works of Byzantine ecclesiastics such as Balsamon and Theodore (Soulis, 1961: 141-2). It is

possible that this group began to migrate westward as a consequence of the earlier destruction of

Armenian cities in eastern Anatolia such as Artsn in 498 AE/1049 CE, arriving in Constantinople in

the early 1050's CE. The story of the Adsincani, according to the Life of Saint George the Athonite

of Iviron (1062 CE; see Peeters, 1922: 36-37), might seem to confirm this arrival date, but again,

caution is necessary identifying this group with those later Egyptians, mentioned in the sources (Fraser, 1992,35; 45). These would be the small-scale traders and purveyors of goods that I

discussed above, and from this period following a different route through Byzantium and south-

eastern Europe, into central and western Europe, than the "dukes" and their retinues.

The narrative I would suggest is one of a tri-partite division of the proto-Romani group from post- Dandängän Trans-Oxiana, with one group maintaining a military function associated with military

189


leadership under Armenian doux and komes. A second group (or perhaps groups) make their way

through Anatolia to Constantinople, and were made up of people with no military function,

probably from the auxiliary personnel associated with the warrior group. These adopted commercial

nomadism and purveyed small crafts and services to survive. Finally, a third group emerged from

the chaos of eastern Anatolia in this period, who were cut off by Saldjük incursions, and migrated

into Karabagh, Georgia and the Caucasus, where they developed a distinct lexicon that evolved into

Lomavren. This schematic description is an attempt to contextualise the complex factors involved in

the emergence of Romani identities in eleventh century Anatolia, and explain the variety and

differentiation that the historical record begins to present us with from this period. It is of course

possible that the survivors from Dandängän were dispersed into small bands, travelling along the

silk routes as "clans" that had varying compositions, in terms of personnel (predominantly military

or auxiliary), with different trajectories through eastern Anatolia into Byzantium, Cilicia or Georgia.

In some senses, this more complicated view could offer a better solution, as notion of various

groups that making their way westwards would underpin the current linguistic view of a number of

successive migrations into Europe, discernible in "three distinct historical layers of Indo-Aryan

innovations" (Windfuhr, 2003: 415). As discussed earlier, it may also be the case that the movement

of those groups speaking what became Lomavren might have been entirely separate to the those we

know as Romani speakers (Windfuhr, 2003: 415; Fraser, 1992: 41).

The over-emphatic application of linguistic models to the problem of ethnogenesis has provided a

framework that the historical evidence has been required to fit, as I suggested in a previous section,

imposing a unitary evolutionary model upon the historical reality (see Hancock, 2005 for a

discussion of the "single migration book", 3-20). In the context of mediaeval Anatolia at the time of

Byzantine collapse and Saldjük conquest, the possibility of differing groups of dispossessed, Indic

dialect-speaking peoples in proximity to each other exists. Descendants of the earlier Ummayad

mawäli troops, the Indian Oigäniyya units, may have been still present on the borderlands of south-

eastern Anatolia and northern Syria. This contested territory had been the scene of continuing

conflict between Arabs and Byzantines during the seventh to ninth centuries, where there is a

documented presence of Zotti or Jhäts (Kenrick, 2004: 32). Possibly descended from groups related

to the defeat and deportation to this region of the rebellious community from what are now the Al

Amarah marshlands of the southern Tigris (Dicle, in Turkish) valley in 834 CE (Kenrick, 2004: 36),

this region still reflects a complex picture of co-existent Domari-speaking and Romani-speaking

Gypsies in the Diyrabakir region. A perspective that suggests the Romani-speakers arrived later

190


than the Dom, but that the two remained distinct (though indistinguishable by non-Gypsies, as is the

case today), might be suggested in that the modem presence of differing linguistic groups may date

from this relatively early period. Patterns of migration elsewhere (Roseman, 1971: 589-98) suggests

that a situation where proto-Gypsy groups arrived in areas where there was an already established

community of a similar kind is possible, although the assumption that each community followed on

directly in a continuous or `chain' relationship is not necessarily one that I would suggest. The

complexity of the actual mechanisms of migration and relationships between differing groups,

speaking related languagess that later emerge as modem Rom, Dom and Lom in Turkey, is beyond

the purview of my current book to examine, but such questions as arise from these possibilities are

possibilities for future study, I think.

The emergence of Romani identities in the chaos of late 11`h century Anatolia, during the collapse of

Byzantine authority, Roman-Hellenic culture and the destruction of the Armenian kingdoms, is the

pivotal point in the narrative of the Gypsy peoples. This is the historical moment when the

coalescence of various elements from refugee Hindü and Khoräsänian soldiers, Persians,

Adharbäydjänis, Georgians, Armenians, Greeks and others came together to forge a new entity. The

shattering of that emerging entity in its very formative process in the confrontation with Saldjükid

expansion meant that differing trajectories were followed, and that for one group, the future

development of their language went in a fundamentally different direction, largely uninfluenced by

the long sojourn in the Greek-speaking lands. For those groups that maintained a military function

under leadership of Armenian counts and dukes (comes and dux, both military designation in

origin), the direction of movement took them into the Cilician territories of Lesser Armenia, until

the destruction of that rump-state and its incorporation into the Mamlük Sultanate of Egypt after the

Battle of Mari (1266; Stewart, 2001: 33-4), and then the growing power of the early Ottoman

Empire.

The Byzantine Empire was the next point of encounter for these groups of proto-Gyspy people who had not moved south after the defeat of the Armenians and the end of the Bagratid Kingdom proper in 1064 at Ani, as they fled before the advancing Turkmen. The experience was the most important

in the process of this coalescing, emerging identity as it was here that the confluence of their arrival

and the notions associated with magic and sorcery came together to `create' or construct the `Egyptians'. The Empire was a complex and multi-ethnic polity that continued as a direct

descendant of the Roman Empire in the East; indeed the term Byzantine is only a modem one that

was never used by the Romioi themselves (Baynes and Moss, 1949: xxiv). The Empire was Roman

191


and Christian, the Oecumene as it was conceived of grew from this centre (Obolensky, 1971). The

history of the Empire is no longer regarded as the "monotonous story of the intrigues of priests,

eunuchs, and women, of poisonings, of conspiracies, of uniform ingratitude, of perpetual

fratricides... " nor is the Empire considered "... the most thoroughly base and despicable form that

civilisation has yet assumed... " (Lecky, 1870: 13-15). The history of the Empire has been

thoroughly rehabilitated from these kind of damning judgements, however it is beyond my purview

here but clearly its central place in the formation of Gypsy identity and the impact of Byzantine

Greek upon the emerging language of the Gypsies means that I would like to consider aspects of the

East Roman state that may have been less well-investigated in the context of Romani Studies

previously. There are two areas of consideration I wish to undertake here; the question of the

Gypsies and taxation in Byzantium, and that of magic and sorcery.

192


6.2 Taxation, Gypsies and status in the Byzantine Empire: the origins of the `Egyptians'

The Byzantine Empire from the 8th to the 11th century had seen what Donald Treadgold suggests

was a remarkable political and cultural revival, a recovery from the profound problems of the period

600 - 700 CE (1979: 1245). The huge loss of territories to the advance of Islam in the east and the

expanding Bulgarian Empire in the west (Treadgold, 1979: 1246), had also caused a profound

reduction in population and consequently revenues from taxation (and military forces). Revenues

had also been reduced through the disruption to trade routes that Byzantine emperors had

previously enjoyed the benefits of, as they ran through the Empire's territories. By the 11th century

this had been reversed. During the century, some important changes occurred that had profound

implications for the Empire and may even have weakened it once more, leading to the conquest by

the Latins in 1204 (Hussey, 1950: 71-2). The confrontation between the state, on behalf of the small

landholders and farmers (groups that had traditionally formed the majority of the military and naval

forces, through these hereditary classes) was won by the magnates, who proceeded to establish the

very large estates that were to become a feature of Byzantine society. The tax burden that had

previously been firmly placed upon the rich was shifted to the poor and immunities were

increasingly granted to the great magnates and land-holders, both lords spiritual and temporal. Tax

revenues were further undermined by the loss of control over the corn monopoly in the middle of

the century to the elite who could exercise patronage through its manipulation. The gold standard

was destabilised in the last quarter of the century, deeply affecting Byzantine trade (Hussey,

1950: 71-75). The irruption of the Saldjüks into Asia Minor and its transformation into Anatolia was

just one of the aspects of the decline in the east; the Dani§mends established control in the north

eastern regions after the Saldjük defeat of the Armenians in 1064 and the Byzantines in 1071 CE

(Vryonis, 1971: 48). Other Turkmen dynasties such as the Karamänids also began to establish their

power bases in a constellation around the foundation of the Sultanate of Rnm, whilst the rise of

Lesser Armenia, or Armenian Cilicia under a cadet branch of the Bagratids further reduced the

Empire's lands in Anatolia. Al of this was to have a major impact upon the revenues of the Empire,

and result in increasingly desperate measures to raise taxes on the part of the bureaucracy.

The question of taxation of Gypsy populations in Byzantium (or more broadly in the Balkan lands),

has been one that has been discussed in the literature (see Marushiakova & Popov, 2001: 16,17,18;

White, 1997; Speck, 1997: 37-5 1; Fraser 1992: 45-9,80,82,157,223; Soulis, 1959: 154-156;

Panaitescu, 1941: 58-72 for examples). It has also been present in discussions of taxation of Gypsies in the Ottoman Empire (Fraser, 1992: 175-176; Marushiakova & Popov, 2000; $imýirgil,

193


2001: 266), most often as a basis for concluding that Ottoman officials (and by extension the

Ottoman state) deliberately discriminated against Romani groups as Gypsies (Ginio, 2004: 118), but

much less is known about the taxation status of Gypsies in Byzantium, as there is little surviving

documentation. The taxation system itself has been the subject of much discussion during the period

of the Byzantine Empire's existence (see Kazhdan, 1991: 377-8; Setton, 1953: 225-59; Sabatino

Lopez, 1951: 209-34; Andreades, 1948: 81; Sabatino Lopez, 1945: 1-42; Katz, 1938: 27-39;

Ashburner, 1915: 76-84), though the fiscal system as it related to the period following Emperor

Diocletian's reforms remains unclear in many respects (Kazhdan, 1991: 377). The need here is for a

brief survey of some aspects of that system as a means of contrasting what I would suggest is the

situation for the Gypsies, with the conclusion that these provided a model for the later Ottoman

taxation system and the treatment of Gypsies in particular, somewhat challenging Ginio's argument

that the differentiation in respect of the Gypsies during the Ottoman period was entirely based upon

their perceived ethnicity. It was in effect an extension of earlier Byzantine practices, not specifically

developed for Gypsies (and not applied to all Gypsies, as the many tax exemptions for particular

occupational or service groups indicate; see Malcolm, 1996: 115,116 for examples), and in line with

the policies of the Ottoman state towards groups that didn't easily fit into the standard system.

Lidner has argued that nomadic (Yörük and Turkmen, wandering dervishes and holy men or

hocalar) groups of all kinds faced increasing mistrust and discrimination on the part of the

Ottomans, as the state became more cohesive and less based upon the tribal and ghazi elements

(Lindner, 1983: 22). Taxation in this context was both the means to control and restrict nomadic

groups (through enforcing a tribal levy annually to extract taxes), and encourage sedentarism

(Lidner, 1983: 23-4). However, the operation of taxation collection in the Byzantine Empire did not

deal especially efficiently with the numbers of nomadic groups that existed in the imperium

(Kazhdan, 1991: 378) other than the Gypsies and then perhaps only those that fell within the control

of state and ecclesiastical authorities in Constantinople.

One of the arguments I make in this book that the taxation status of the Gypsies in the Byzantine

(and later Ottoman) Empire was a significant indicator of their social and economic position but

not, as has so often been stated, entirely a consequence of their ethnicity. In this context, I suggest

that the combination of late Byzantine and Ottoman taxation systems placed the Empire's Gypsies

in a very particular position with a particular status. The key point that is to be remembered is that

taxable populations are those that have social and economic value in the context of the wider

society here. Non-taxable groups of the Byzantine population were usually those at either extremes

194


of the social hiearchy, i. e. slaves, colonii, the elderly, women, children, or those granted tax-

exemption on the basis of their status as members of the religious or military classes. The Byzantine

taxation system was effective at extracting maximum surpluses from both indirect levies upon the

rural population and direct revenues from customs and excise duties, retail dues, transit and port

taxes and a bewildering variety of supplementary and extraordinary imposts especially during times

of war (Andreades, 1948: 83). The extent of the Emperors' incomes from urban resources was

substantial: all caravansaries and bazaars belonged to the state rather than being ceded to the

merchants gratis for their use. Mines and quarries were equally part of the state fisc, and the

revenues from customs duties and consumption could provide the Emperor with a sum of 20,000

gold pieces (nomisma or bezants) a day at the Empire's height. One of the most important trade

goods that the Emperor controlled directly was silk, and the purple dye and gold embroidery used to

embellish it, that formed not only a rare and precious commodity, but was an aspect of political

power in addition, as Sabatino Lopez has shown (1945: 1-42). The prominence of specie in the

Byzantine Empire was such that, despite the mines and quarries, salt-pans and customs, together

with the direct levies upon the agrarian population, the gold nomismata coin was used to pay most

taxes (Sabatino Lopez, 1951: 209-34). The major difference in taxes was based upon an

urban/agrarian duality; where markets, ports or transit routes existed there were no hearth or

capitation taxes (Andreades, 1948: 81). The other major source of public revenue in the Empire were

the State properties; those industrial manufactories, agricultural estates and urban properties

mentioned above (bazaars, caravansaries) owned and managed by the imperium. These provided

foodstuffs for the Court, the bureaucracy and victualed the military in the capital and administrative

centres. Arms were manufactured, including the mysterious Greek fire, in some industrial properties

and the production of luxury goods both for the members of the Imperial household and as

emoluments and gifts for distributing largesse was undertaken by others (Andreades, 1948: 84). This

fundamental distinction made in Byzantine taxation was between liabilities upon land -jugatio -

and those upon human beings and animals - capitatio - in what might be described as hearth and

poll taxes. The relationship between those that cultivated the land and the land itself was strictly

controlled and subject to supervision by the state officials - it was not permitted for peasantry to

leave or be removed and taxes were hereditary, though chiefly in kind. These had to be transported

and distributed through state control.

One interesting point concerning the Gypsies might be connected to this aspect of Byzantine

taxation; the division between those who were bound by law and duty to the land as peasantry

195


paying taxes and dues to the land-owners (Birkenmeier, 2002: 153; Kazhdan, 1991: 2015), and those

who were not, or attempted to define themselves as independent of this system. The fundamental

divisions that had existed during the 7'h to the l Ot" century between those who inhabited the landed

estates of the aristocracy, the colonii and the paroikoi who were largely dependent serfs, and the

georgioi peasant farmers who paid taxes that supported the stratiotai peasant-soldier class, had

become much undermined by the 11`h century (Charanis, 1953: 415). The aristocracy had

successfully defeated the imperial attempts to limit their acquisition of landed estates, particularly

through the mechanism of the award of pronoiae, originally intended as a land grant to soldiers to

recuperate deserted lands and reinvigorate the army. However, these became invariably awarded to

men of high estate, not the soldier-peasantry of earlier periods, and despite the prohibition of them

becoming hereditary, by the 13th century they had become incorporated into a system of appanages

amongst the great families or dynatoi (Charanis, 1953: 419; Birkenmeier, 2002: 148-154). Certainly,

at the time of the Battle of Manzikert (1071 CE), the extent to which the traditional soldier

peasantry core of the Byzantine army had been undermined was apparent to Georgios Kedrenos,

when he wrote:

"The army was composed of Makedonians and Bulgarians and Varangians [Vikings from

Kievan Rus'] and other barbarians who happened to be about... The renowned champions of

the Romans [Byzantines] who had reduced into subjection all of the east and the west now

numbered only a few [amongst these]... and these were bowed down by poverty and ill-

treatment. They lacked in weapons, swords and other arms... " (1839,11: 652)

Clearly, the Byzantine armies of the later period were poly-ethnic, and multi-lingual but did not

include the `Egyptians', which strongly suggests the complete loss of any military functions

amongst the residue population present in Constinople and elsewhere in the Empire, concomitant

with my earlier argument that the warrior class had been decimated in the confrontation with the

Saldjükids at Dandängän and Ani. Had such remained, it seems likely that the Byzantines would have enrolled these elements into their fighting force along with Turks, Alans, Ossetians, Anglo-

Saxons, Normans, Flemish knights, Peeenegs and others (see Birkenmeier, 2002: 206-30; Treadgold,

1998: 85; Bartusis, 1996). Yet as we know, there were bands of soldiers enlisted under Roma

command by the Venetians in their campaigns against the Ottomans, defending their Greek

territories in the 1440's (Fraser, 1992: 51). This gives us a further indication of the status of `Egyptians' in the Empire, as neither taxable with the peasantry nor employable (or conscriptable)

with the military.

196


I would suggest here that the debate around the term gadj6/gadje and its etymology, that Hancock

defines as "civilian" (2002: 10), and that Matras argues is mistakenly glossed in order to support the

warrior-origin book (Matras, 2002: 205), is pertinent. Matras argues that there is no solid etymology

for the word, though one author derived it from the Old Indo-Aryan word garhya, meaning

`domestic' (2002: 205). This would appear to suggest that the ancestors of the Gypsies, or in this

case the Egyptians of Byzantium, had an arguably negative concept of `domestic', which seems

highly unlikely as these people were not nomadic in their origins, nor during their sojourn in

Ghaznä and surely even nomads have conceptions of domesticity, as applied to their own living

spaces, however mobile. The more plausible origins I think lie in this Byzantine distinction between

those who were tied to the land, un-free and obligated to pay taxes and perform services for the

magnates and land-owners who owned the vast estates upon which they lived and worked, and

those who were not part of this system and who would appear to have avoided paying taxes

whenever possible. The latter point might be deduced from the various edicts indicating that

extraordinary taxes should be collected from them (Marushiakova & Popv, 2001: 16), surely unusual

in a highly regulated system where taxes and imposts were widely understood and levied. That such

instructions indicate "from this time onwards Egyptians begin to be included in the category of the

villeins (parici [paroikoi? ]), the dependent feudal population in Byzantium. " (Marushiakova &

Popov, 2001: 16). I would argue here that these "special taxes" (2001: 16) are precisely that, part of

the attempt of the Byzantine state to extract revenues from a problematic population that remained

outside of the system, and in this distinguished themselves from those within. The attempt of

Byzantine churchmen and administrators to include them with the enserfed peasantry in terms of

taxation status was resisted in the development of the language, at the basis perhaps of the later

assertions to be Romioi or Romiti (Fraser, 1992: 53), distinct from the paroikoi, georgoi and

stratoikai.

The state also made taxes upon some occupations hereditary, such as ship-owners, bakers and pork-

dealers, but many occupations were controlled in terms of the organisation of the guild system and

the state encouraged the transmission of artisan craft skills within families. The guilds functioned to

facilitate commerce and provide services as required by the state, whilst the state regulated prices

and quality. Frequently the guilds also provided labour for communal projects such as repairing defensive walls or constructing aqueducts, on a rotating basis. There seems to have been no incorporation of the Egyptians of the capital into these guilds, as happened in the Ottoman case. By

the late Byzantine period the shift from an agricultural to an urban economy had taken place,

197


largely as a result of losses in territory and the concentration of land in the hands of magnates,

creating an effectively enserfed peasantry or paroikoi. The major land owning class also came to

include ecclesiastics controlling huge estates belonging to the monasteries, many of which were

exempt from direct taxation. However, despite the attempts by both urban authorities and the

Orthodox Church, the Egyptians appear to have existed as a group that attempted to remain beyond

the control of the state and its agents, most especially the clerics and only indirectly paying taxes

when these were specifically directed at them. That the state attempted to do so in the form of a

differential tax, the kaphalition also levied upon other non-Christians (Muslims, Jews, heterodox

Christians, Mazdians, Zoroastrians and others), illustrates that the Byzantine perception of these

groups was primarily as being confessionally different (Andreades, 1948: 82). This was the basis for

the continued differential treatment by the Ottomans, as I suggested above. Their recorded

occupations are most frequently concerned with magic and sorcery, divination and charms, and it is

the question of their relationship to the notions of magic and their identity that I consider next.

198


6.3 Magic and Romitoi in Byzantium; the construction of `Gypsy' identity

In attempting to describe a trajectory for these Gypsy groups moving rapidly through Anatolia

during the eleventh-century (1040-1071 CE), in a previous section, I suggested that the

establishment of the Romitoi identity as a phenomenon, took place in the latter half of the eleventh

century CE. As these elements coalesced that had previously been mawäli or client soldiers and

their auxiliaries in the Ghazndvid armies, others became added to these that were from the

Armenian kingdoms, defeated by the Saldjüks after 1064 CE, and refugees from Turkmen

incursions. The mechanism for the shift from these groups of former military function, through

operating barter and exchange for basic goods and services as refugees after the defeat at

Danddngdn (1040 CE), to peripatetic commercial nomads exchanging goods and services for cash,

is to be found in the seminal contact between Romitoi and urban Byzantium. This contact

fundamentally altered the nature of these communities and re-defined them as a distinct

occupational groups with an ascribed, pre existent identities (Egyptians/Atsinganoi), servicing the

urban markets of Byzantium, especially Constantinople, and altering their earlier relationship to

political structures from client soldiers to service nomads. The impact of Byzantine technology and

economics is to be measured by the importation into the Romani lexicon of large numbers of loan-

words associated with metal-working, counting and money. Reduced from skilled artisans

producing high-quality armour and weaponry, using the resources and materials available to the

highly centralised Ghaznävid military, proto-Gypsies (if we describe thus the groups moving

through Khorasan and Adharbaydjäni territory, speaking a language derived from the Ghaznävid

koine) now worked with basic, portable tools and materials to produce low technology implements,

probably for rural inhabitants and town-dwellers. Contact with the more sophisticated markets and

technology of Byzantium produced a further profound change in language and, I would suggest, in

the structure of economic relations and in the context of the stress and turmoil of conflict and chaos. This picture is to be seen in the context of the other process I outlined previously, of military

contingents associated with displaced and dispossessed nobles able to establish petty principalities in the Cilician and Taurus regions that became known as Lesser Armenia. Their trajectory is more difficult to follow, and more speculative in that the existing sources give few indications of what this may have been, until later in the 15th century. It might be hoped that future research may indeed

qualify or further refine the picture of the twelfth and thirteenth centuries in these lands through further investigation of documentary sources in the Turkish and Armenian lands. Until then, my suggestions must remain a constructed narrative around very few fragments drawn from linguistics,

199


however plausibly it may offer an explanation for the historical circumstances. I shall return briefly

to this when examining the arrival of the Gypsies in fifteenth century western Europe, but here will

attempt to define more closely the picture of Egyptians in Byzantium.

My argument in this case, is that Gypsy identity was initially associated with notions of magic, as

far as the Byzantines understood there to be a connection with `magical' Egypt, and a variety of

Romitoi groups practised fortune-telling, sorcery and divination, whilst others occupied less exotic

commercial niches (blacksmiths, cauldron-makers and farmers; see Marushiakova & Popov,

2001: 18). The classical and Hellenic heritage of locating `fabulous peoples' in India and Egypt,

through the prism of Byzantine commentators like Procopius and descriptions of `barbarians'

surrounding the Empire, pre-disposed 11th century Byzantines, especially amongst the intellectuals,

to this view (Kartlunen, 1989: 127). The initial identification of Gypsies was through a conflation

with a group Judeo-Christian `heretics' from eastern Asia Minor, known as Atsinganoi (Fraser,

1992: 46). The question that arises is to what extent this can be seen to be justifiable, and Fraser has

correctly questioned this point as being far too frequently resorted to by Romani Studies scholars

seeking evidence of Gypsies in Byzantium from a very early period (Fraser, 1992: 46). As he

remarks about much that passes for "evidence" of Romani presence in various contexts, "such

theories are not infrequently seasoned with a liberal measure of subjectiveness" (1992: 32). The

mention of Atsinganoi in a hagiography of St. Atanasia, records her giving food to foreigners

described as such, in the year 800 CE, and is cited as evidence in an account of the origins of the

Gypsies, drawn from a wide number of secondary sources. This account goes on to give the various

attributed references to Atsinganoi, and explicitly assumes a direct link between the two. Apart from

the mistake of the saint involved (St. Atanathios, rather than Atanasia; see Mango and Scott, 1997:

473), the text is representative of much of the conventional wisdom amongst some writers about

early Romani history (see Marushiakova & Popov, 2001: 11-18). More important, it embodies the

almost universal desire amongst them to establish a presence in Byzantium at an early date, and so

fit in with the time-scale suggested by accepting the Harriott book regarding the description of the

La in Firdawsi s Shäh-näma, that I analysed in an earlier section. Thus a 5th century presence in

Persia would allow for the presence of Romani people in Byzantium at the opening of the ninth

century CE. The linguistic evidence clearly does not support such a view, and this book has argued

that this trajectory, or "narrative of journey", is fundamentally incorrect.

The question of the heretical group from the Phrygian region called Atsinganoi, remains important,

despite, or perhaps because of, the obvious conflation with the Egyptians in Byzantium. Hamilton

200


and Hamilton have suggested that the Athingani (from which the former term is derived; see Fraser,

1995: 46) made no secret of their beliefs, or attempted to make converts, and so were not deemed to

be a threat to the Church (1998: 69). Euthymius of the Peribleton, writing in about 1050 CE,

suggested that the heresy of the Atsinganoi, unlike that of the Bogomils, was both inherited and

obvious, and not able to harm or upset anybody (Stoyanov, 1994: 139). Another factor is to be

found in the association of ideas of divination and astrology practised by both the heretics and the

later Gypsies, and the possibility exists of early groups describing themselves as from `the East',

thereby encouraging a false correlation in the minds of those who first met with them. Byzantine

nomenclature was notoriously archaic in its ascription of ethnic labels, frequently resorting to

classical writings to describe groups such as the Huns as ̀ Scythians' (Hoffman, 1973: 1350), and in

this context the use of the term `Atsinganoi' for these new `semi-barbarians' from the east is

perfectly consistent. There also exists the possibility of some, though not all of these proto-Romitoi

being in contact with communities related in the past to the ninth-century Atsinganoi. Their

trajectory through Persia may have brought them into contact with Zoroastrians, Manicheaens and

other dualists; Asia Minor was rife with `heretics' like Bogomils and Paulicians throughout this

period of crisis, but it is significant that the charge of heresy was never made. Additionally, the

influence of central Asian shamanist beliefs and practices, together with the emerging Anatolian

Turkish culture may also have affected Gypsies, though only one reference is to be found in the

story of the Atsinganoi recorded in the hagiography of Saint George the Athonite (Peeters,

1922: 102-4). The notion of them being "a Sarmatian people" might suggest elements of shamanism,

though it may equally be a Byzantine influence in this Georgian text (Fraser, 1992: 46). The term

continued to be used concurrently with Egyptians in descriptions of these groups, suggesting that

there may have been internal differentiation that was not clearly understood by the Byzantines. The

Patriarch of the Holy City, Gregorios Kyprios (1283-9) mentioned both in an ordinance relating to

the collection of taxes (as mentioned in the previous section; Marushiakova & Popov, 2001: 16), but

the term Egyptian seems to have become more widespread at about the same time (Fraser, 1992: 47-

8) in similar contexts. The additional reference to Egyptians as acrobats might also indicate some difference in perceptions, in the case of the acrobats who arrived in Constantinople around 1314 CE

and interviewed by Nikephoros Gregoras (Soulis, 1961: 148-9), though it is also possible that these

were Dom and not Rom, as they did indeed hale from Egypt. However, the connection with Egypt

and magic was an important part of Byzantine culture at all levels from the l lth century onwards, as I shall show.

201


The question of the relative importance of magic in mediaeval Byzantium has been addressed by

Maguire et al (1995), and has a direct bearing upon the history of the Gypsies in this context. The

periodic attention that magic and magical practices received in Byzantine scholarly and

ecclesiastical writings is an indication of the interest, both academic and popular, that the subject

received (Maguire, 1995: 10). Magical practices and sorcery had been indistinguishable and poorly

defined in relation to Orthodox Christianity, despite the successive attempts by the early Church

Councils to define and legislate against them. The concern with magic in Byzantium was first

apparent in the 4`r' century CE, when it became the object of secular law (Fögen, 1995: 99). The

study of magic continued throughout the Byzantine period, to a greater or lesser extent, until the fall

of the Empire, and Greenfield has demonstrated that Paleologian magic was flourishing until the

last days of Byzantium, perhaps understandable in the context of the collapse of the imperium and

the daily insecurity of life in the empire (1995: 117-153). Divination from weather, clouds, stars,

portents and even the markings on the shoulder blades of sheep, were common practises, in addition

to the bear-leaders, amulet-sellers, soothsayers, palm-readers and `scrying-stone' diviners (crystal

ball readers) (Rosser, 1997: 859). The question of definitions, both internal and external, during this

time-scale is of course important; what was understood to be magic at any point in time varied, as

did what was acceptable in the context of Orthodoxy, and what was "illicit" (Maguire, 1995: 3).

Curiously, the infamous malleus maleficarum, so central to the European Dominican Inquisition,

never appeared in Byzantium (Rosser, 1997: 859). As Fögen amply illustrates, the significant

change came in the 12th century, when the "domestication" of magic resulted from the greater

precision in defining what constituted dealings with the supernatural, and altered these from being

the province of secular law, and draconian punishment, to cannon law, and the prerogative of the

church, who came to see this as a question of indiscipline, rather than treason (1995: 99-104).

The initial reaction of secular and imperial authority to magical practices must be seen in the

context of the challenges arising from heresies that, at points, undermined the authority of the

Orthodox Church and the Emperor, as its head (Hamilton & Hamilton, 1998: 12-17). The great

Church Councils of Nicea (325 CE), Constantinople (381 CE) that had condemned the Arian heresy

left unresolved by Nicea, and further clarified the position of the Holy Spirit by determining that it

was "the Lord, the Giver of Life who proceeds from the Father... " thereby establishing the doctrine

of the Trinity (Hughes, 1961: chap. 2). The Council also approved the primacy of Constantinople

(New Rome) over all other sees, after Rome, though this was never accepted by the Apostolic See

and the papal legates refuted this at the Council of Chalcedon (November 451 CE). The Council of

202


Ephesus was also called to deal with the controversy surrounding Nestorianism (431 CE) and

Chalcedon to counter the growth of further Nestorianism and the teachings of Eutychus. Attempting

clarify the decisions taken at Ephesus in 449 CE, at the so-called "robber council" (Hughes,

1961: chap. 4), the Council became embroiled in disputes over the formula of adherence to the

doctrine of the Incarnation of Christ, thus precipitating the split between "Oriental Christianity" or

"Orthodoxy" (Coptic, Syriac, Armenian and Ethiopian) and the remainder of the Church. These

Councils were not concerned with magic so much as heresy, though they eventually came to

consider magic and its practices with the Council of Trullo in 691-2 CE. The first Canon of the

Council reiterated the findings of the Councils of Nicea, Constantinople, Ephesus, Chalcedon, the

Second and Third Councils at Constantinople regarding doctrine, before proceeding to examine

various matters including magic. The Council itself was not fully recognised by the Latin church in

the west, as the Pope refused to sign the canons meaning that they were never observed in that part

of the Church. Justinian II's aspirations that this should have been a truly oecumenical council were

never realised, as even in the east, the Trullo canons were treated as less than authoritative

(Schaffer, 2005: 503).

By the time of Balsomon's commentary upon the canons drawn up at Trullo, the attitude of the

Orthodox Church to notions of evil had been further clarified from the 4`h century conceptions of

the Church Fathers (Dickie, 1995: 9-10). In some ways, the figure of Satan had come into focus as a

result of deliberations regarding the nature of evil and the relationship between Satan and God

(Stewart, 1991: 113,189). The refutation of any notion of dualism, in that Satan might be considered

an equal power had established him as part of God's Creation and therefore subject to Divine Will.

Satan in fact had no independent power and any success he had in corrupting mean and women was

entirely due to their laxity or weakness, lapses of the determination to live according to the tenets of

God. Satan was immaterial and not to be associated with particular locations, with very limited

abilities to tempt or delude temporarily, in a variety of forms that were largely indistinguishable

from him and certainly all of the same order. Importantly, the notion of a host of intermediaries was

absent, there appeared no complex taxonomy of daimones or demons such as came to feature in the

later Roman Catholic conceptions; such things were regarded as outside of the Christian order and

dismissed. As such, the existence of ambiguous xotikd, or the Hellenic corpus of spirits, sprites and

demons was not accepted and regarded as ignorant superstition amongst the masses (Lawson,

1910: 130-41). In actual fact the notions of two opposing forces in the world, what Stewart calls "a

world view [that]... is predicated on the existence of at least two opposing sides" (1991: 99), would

203


seem to have existed in the Byzantine world, as a relic of the pre-Christian Hellenic beliefs and the

influence of shamanism from the Pontic regions (Dodd, 1951: 267-9), certainly in popular culture.

The 11`x' century saw a scholarly revival of interest in magic, alchemy and astrology as an academic

concern, particularly by intellectuals such as Michael Psellos. Duffy suggests that Psellos was

instrumental in reintroducing major works on magic and magical practices that had hitherto been

unregarded (Duffy, 1995: 83-4). This concern with what was clearly in opposition to the teachings of

the Church and Psellos' own position, was surprising, though Psellos himself seems ambiguous in

his repudiation of the works he studied as pagan nonsense and Michael Italikos is contemptuous

throughout (Duffy, 1995: 86). The complex attitudes towards magic shown by some Byzantine

intellectuals was however, not reflected in popular beliefs about the efficacy of magical practices, or

what might be described as witchcraft as opposed to sorcery - that branch of magic pursued by

scholars and intellectuals like Psellos and Michael Italikos (Greenfield, 1995: 119; Duffy, 1995: 91 -

97). In popular belief, the existence of demons, faeries and all kinds of xotikd was widely, almost

universally accepted (Stewart, 1991: 101).

One other factor needs to be further mentioned here (as referred to briefly above), the question of

dualism and more particularly Bogomolism. The 11th century saw the rise of the Bogomols in the

Byzantine Empire, a branch of what Runciman has described as the mediaeval Manichee (1947).

The foundations of Bogomolism in the 10th century appear to be based in earlier Pauliciansim and

Massalianism, according to Anna Comnena (Dawes, 1918: 412-415). There has not been any

connection drawn between the Gypsies and Bogomolism and it's not my intention to do so here, but

to suggest the context of Byzantine society into which the `Egyptians' arrived (for histories of the

Bogomil movement see Fine, 1983: 171-9; Hamilton & Hamilton, 1998; Stoyanov, 1994: 191-207;

Obolensky, 1948). No Byzantine documents assert the charge of Bogomilism or even heresy about

the Egyptians, which in itself is perhaps surprising given the possibility that the proto-Gypsies are

likely to have been in contact with Christianity in its Armenian miaphysite variation, as Hancock

has shown (2002: 74). This would suggest that the use of the notion of the Armenians as heretics

was largely political, rather than religious, in the struggle for control of the Bagratid Kingdom in the

11" century. The Gypsies were clearly associated with sorcery, in the earliest reliable documentary

reference to the Atsinganoi (Fraser, 1992: 46), but it would appear that this association was more by

inference than any suggestion they studied magic in the same way that intellectuals in Byzantium

did. The further references note them as soothsayers, charmers and magicians, and more darkly as

"teaching devilish things" (Soulis, 1961: 146-7), but the context that I have outlined above would

204


explain why this was not perceived as socially disruptive or a threat to the oecumene - this was a

product of ignorance and superstition, lapses in will on the part of the gullible but not the actual

work of heretics from within.

The specific commentary of Theodore Balsamon on the edicts of the Council of Trullo, held in the

late 7th century (691-692 CE), is central to aspects of the discussion of Gypsies in Byzantium, and is

frequently referred to in the literature following Soulis (1961: 142-65). The particular canon that is

referred to is sixty-one, where he describes those who are mentioned in more detail:

"Those who expose themselves to soothsaying or to the so-called ekatontarxois [old, or

wise-men who mislead the simple] or similar people, to hear what they wish to be disclosed,

are to be subjected to six years of penance according to the rules of the early fathers. To the

same penance one must submit those who drag a bear or similar animal after themselves for

the enjoyment and the damage of simple-minded people and who tell the future, fate,

horoscope, and whatever else may be the multitude of words of this erroneous trumpery.

The same is true for the interpreters of the clouds, sorcerers, furnishers of amulets, and

soothsayers. We decree that those who continue doing so, who neither show repentance nor

avoid these destructive and pagan customs, shall be totally expelled from the church

according to the holy canons. "For what communion has light with darkness? " [2

Corinthians 6: 14-16]" (Fögen, 1995: 103)

After mentioning the ekatantarxois, Balsamon refers to the bear-leaders, who are not merely

showmen, but dyed their animals, cut strands of their pelt and sold them as amulets against the "evil

eye", a series of shifting notions about malign influences that stemmed originally, from envy

(Dickie, 1995: 12), and other baleful influences (Fögen, 1995: 101). Other magical practitioners are

enumerated, but it would seem the Egyptians as Atsnganoi, are one of the primary concerns for the

churchmen of the Byzantine Empire at this time, as Balsamon goes on to mention them again in

canon sixty-five (Fraser, 1992: 47). This time the connection with evil is more emphatic, as the

Atsinganoi are inspired by Satan to false prophecy, and perhaps somewhat strangely, ventriloquism

(which might indicate another sphere of entertainment that these groups were involved in; Fogen,

1995: 102). Fögen argues that this enumeration and attempt to clearly categorise is a significant shift

from the 4`h century prohibitions against all those who practised magic, and away from secular

punishments to a more accommodating perspective that saw this as within the purview of the

church (1995: 104-5). As lapsed Christians, those who consulted the Egyptians were to be seen as

deserving of mercy and forgiveness after a suitably penitent period (Fraser, 1992: 47).

205


The arrival of the Egyptians coincided with a number of important trends in Byzantine society, as

outlined above. The earliest Church Councils had defined orthodoxy and faith in a way that, whilst

not eliminating heresy altogether made it clearer who might be defined as beyond the oecumene,

through wishing to harm or undermine it from within. In this way, those who practised magic and

sorcery (superstitions) may have been deluded by evil influences temporarily, but ultimately their

efforts were worthless and fruitless. Heretics by contrast threatened the very order of the oecumene

and were to be treated differently, as they sought to subvert the Creation through distorting it by

suggesting that Satan was coeval and in opposition to God as an equal, such as the Bogomils. In

popular terms, the notions that sorcerers and magicians operated in the world of xotikä as an

opposition did retain its hold on the imagination of everyday Byzantines, but as these existed

beyond the bounds of Christian theology (unlike heretics), they could be dismissed or discussed by

intellectuals and theologians in a way that defined them as superstition, not social disruption. The

Egyptians were part of this world of xotikä and could be ultimately dealt with through bringing

those who appealed to them back into the community of the faithful with penitence and prayer,

attempting to neutralise them or their influence. As part of the xotikd, and through the association

with Egypt, the Gypsies were intrinsically linked to notions of enchantment (and probably

encouraged such notions in the interests of commercial advantages over other magicians and

practitioners of divination and sorcery). The "uses of enchantment", to quote Bettleheim (1989)

may well be to

This I think suggests that unlike the Bogomils or other groups, the Egyptians were not persecuted in

an attempt to extirpate them, but to some extent (as Fogen argues; 1995: 115) incorporated into the

round of day-to-day administration of the faithful:

"... the initial excitement and chaos, which the secular power first provoked and then did not

get under control, gradually gave way to a professional handling which ended in a matter of

routine. Canon law and its experts, step-by-step, by description and distinctions, transformed

a home made political confusion into the normality of religious discipline. For magicians,

their clients and judges, the world thereby became more calculable, less complex, and easier

to understand. "

Part of this process of definition and incorporation was to ascribe the proto-Gypsies a place in

society that made sense in the Byzantine worldview; as part of the continuum of popular belief in

xotikä and intellectual study in magic, the people that had arrived from the east were to be

associated with the source of all magic, Egypt. Increasingly the association with the older notions of

206


the Atsinganoi was to give way to a stronger identification with Egypt and magic, as a result of

these changes and trends. The association was one that remained and eventually became the term by

which these people and their descendants would be known, and in this sense the construction of

Gypsy identity can be traced to 11`h century Byzantium and the interest in magic that remained a

feature of its culture until its fall in 1453 CE.

207


7.1 Ottomans and Gypsies "" .. my Lord has created a city... " Haci Bayram Veil, d. 1430

"The Turks are a barbarous people borne to the destruction of cities, arts and learning, have prospered more by our

vices than by their own "virtue ". This public calamity of the world, by barbarous violence, multitudes of men and

obedience to severe discipline had grown great" John Barclay, The Mirrour of Mindes, (1633)

EXAMINING THE OTTOMAN EMPIRE IS, in many ways an opportunity to see a different way of being,

a society based upon precepts other than those we see as inextricable with our own sense of

ourselves. These perceptions of Ottoman society and culture were and are often based upon those a

priori criteria and argumentation constructed in ways that may have enormous impact upon our

analyses and certainly less upon what might be considered as "inner Asian norms" (Togan, 1992).

Observation and investigation may provide startlingly different results if conducted from historical

curiosity and re-evaluation of the sources and literature. It is the case that pre-Enlightenment

investigation of the Ottoman world can be seen as characteristically empirical and historical in a

way that in the 18th century became "... unified, settled, fixed and ... self-evident in describing

Ottoman reality in an ahistorical and rational framework of analysis. " (cirakman 2002: 4) The

fluidity and multiplicity of images and notions of Ottomans as admirable, fearful, sympathetic and

militarily superior that can be seen to have existed in 16th century and early 17th century (see Anon.

1597,1622,1685; Baudier 1633; Busbeq 1694; Giovo 1546; Goffe 1631 a&b; Knolles 1603; M. B.

1660 for examples) became refashioned in the 18th. Exotic and execrable, desirable and detestable,

the image of the Ottomans in European thought, literature and political discourse became the

bedrock of 19`h century notions of the "Sick Man of Europe", the "Lustful and Terrible Turk" and

the archetype of "Oriental despotism" and its necessary corollary, decline. Such tropes allowed Europe to find a remarkably common set of responses to the Ottomans, replacing the earlier Catholic antagonism and Protestant alliances of the 16th and 17`x' centuries, with a rational for

economic penetration and eventual dominance (Clay, 2000: 5-7). The stream of renegadös to the

Ottoman lands in the earlier period became a constant movement of travellers, opportunists and

carpet-baggers convinced of European economic and military superiority and Ottoman-Turkish

imbecility, ineptitude and idleness, all providing a commentary for audiences eager for confirmation

of their prejudices (Davidson, 1990: 57). Octavio Bon's early work on the Sultan's Seraglio, an

example of the empirical and historical approach characterising pre-18th century European curiosity (1996), became Lurid and improbable descriptions of the turgid and deliquescent world of the hammam and harem, almost entirely fictitious in origin and limited in imagination (see Grosrichard, 1998). The concerns with Ottoman origins and society, as a means of explaining (and emulating)

208


Ottoman military successes (see Anon. 1597,1622,1685; M. B. 1660; Baudier, 1633), transformed

into condemnation of the corruptness of Ottoman governance over the Empire's subject Christian

populations (Bacon, 1861) and concerns over the potential for a radical re-alignment of the `balance

of power' should Russia prove successful in its attempts to recapture Tsargrad (Constantinople) for

Orthodoxy (McCarthy, 2001: 28),.

As a multi-religious society made up of a huge variety of differing peoples, the Empire needed

different solutions to the issues of governance, social organisation and control than those of the

developing nation-states in Europe (Braude & Lewis, 1982: 7). Models for a successful polity were

primarily Persian and Turco-Mongolian, with the influence of Byzantium (Kunt, 1995: 14-15)

traceable in the local bureaucratic usages. Originally a group of Oghuz TUrks, found in the 13th

century on the hinterland between Byzantium and the Selcuk Sultinate of Rom, the Ottoman

chronicles (like that of Karamini Mehmet Pasha written during the reign of Mehmet 11), claimed

they had arrived seeking new pastures after being displaced by the Mongols of Chingiz Khan. As

successful irregulars in the Selcuk military forces, the Oghuz and Turcoman tribal leaders began to

coalesce around a certain Ertoghrul in the regions of Ermeni Beli and Domanis in summer and

Kishlak, in the vale of Sörgüt in winter. This effectively controlled the direct trading route and

pasture-lands between Bilecik and Eskeshehir (Lindner, 1983: 20). Greeks, Turks, peasants and

townspeople and even the occasional renegade Byzantine lord began to look to the emirs of Caria,

Aydin, Saruhan and Karasi to provide protection and stability in a region where the population had

been effectively "... forgotten by their distant and powerless capital [Byzantium]... " (Kunt, 1995: 7).

Ertoghrul remains an extremely shadowy and part-mythical figure in the late 13th century, whom

Kara Mehmed has dying after many successful campaigns at the age of 93 (Lindner, 1983: 21). His

son and from 1299 holder of the chieftenship, Osman is however a much clearer and more definite

character. It is he to whom we owe the name of the dynasty and the empire that it ruled for seven

centuries.

The picture of Gypsies in the European Ottoman territories has been recently described in detail

(Marushiakova and Popov, 2001; Seeman, 2002; Celik, 2003) with regard to their social and

economic status, particularly with regard to how Ottoman administrators viewed Gypsies and

imposed differential taxation levels upon them, arguably as a consequence of the perceptions of

them as an ethnicity (Ginio, 2004). That this represents a continuance of late Byzantine taxation

policy (and perceptions of the Gypsies) I have suggested above, and the principles by which such decisions were made have been demonstrated by Co*gel (2004) to explain variations driven by

209


economic considerations, whatever the ideological justification that may have been mobilised as a

consequence (2004: 18-28). Briefly, the costs of measuring the tax base and collecting the revenues

was always considered against the consideration of maintaining local practices and rates of taxation

and adapting these to the needs of the state for efficient revenue collection (2004: 27). Political

assimilation of conquered populations may have been less of a consideration when it came to

groups like the Gypsies, but the process of collection was always adapted to suit local conditions -

in the case of the Gypsies the annual festival of Hdrilez (5th and 6th May) in Thrace was the

opportunity to collect taxes from nomadic Gypsy groups, hence the term by which the Turkish

Gypsies still know it as Kakava, relating to the copper cauldrons used by families for cooking food

and the basic measurement of units of taxation by the Ottoman enumerators and tax collectors (this

explanation was provided by various Romanlar Gypsies during fieldwork in Edirne, May 2005).

The notion that ethnicity was noted in legal cases amongst Gypsies in Salonica, in contravention of

normal Ottoman practices (Ginio, 2004: 118) is misleading, in that the regulation of Ottoman society

was always based upon complex ethnic, religious and class distinctions, as Makdisi has argued

(2002: 768). In counting population, ethnic differences were indeed recorded as Karpat has

documented (1985: 115-221), including Jews, Armenians, Bulgarians, Greeks, Yörüks, Alevis and

Tahtaci all noted in the population count of 1831 (Karpat, 1985: 111-115). Clearly `normal' Ottoman

practice in a variety of administrative contexts did note the ethnicity of individuals, as well as their

confessional adherence.

The picture of Ottoman Gypsies may be added to here by a narrative drawing upon a different

source than that usually resorted to in discussions of Gypsies in the Ottoman lands, that of the

Turkish Gypsy communities themselves, and in particular that of the Istanbul neighbourhood of

Sulukule, the area where the Gypsies in Byzantium are first recorded in 1050, as eradicating the

wild animals in the royal hunting ground of the Philopation, just outside the land walls to the west

of St. Saviour in Chora monastery (see Sevcenko, 2002: 74). The Ottoman period is an important

point for the cultural heritage of the Sulukule neighbourhood in Istanbul as the community maintain

a strong tradition that this area was given to them after the taking of the city in May 1453; the

Conqueror Mehmed II was supposed to have awarded this for the efforts of the Gypsy troops (they

were most frequently commanded by Crimean Tartars in the Ottoman period, whilst the Gypsy

canon-founders and blacksmiths were under the command of Hungarian engineers) in their first

attack against the Azi Kapi, or Gate of the Assault (now on the edge of the Sulukule quarter), clearly

using them as ̀ shock-troops' to be followed by the attack of the janissaries. 6,000 of them are

210


supposed to have died in the assault upon Constantinople, and are described as ̀ martyrs' in the

narrative. This tradition provides the community with its focal point of a claim to historical

occupation since 1453, though by the population counts undertaken during the early Ottoman

period in Constantinople/Istanbul, the numbers that resided there were at first quite small (a 1477

count suggests some 41 households). The re-population of the city in the twenty-five years

following the conquest brought large numbers of Gypsies from Anatolia (places like Balat) to the

area and increased the population rapidly in the neighbourhood. The position of Gypsies in the

military organisation of the Ottomans, as metal-workers (especially in the Tophane complex),

drovers, grooms and horse-trainers, porters, powder-makers, fletchers (arrow-makers), tent-makers

and a host of auxiliary roles was an essential one in the huge machinery that prosecuted the

campaigns of conquest engaged in by the Ottoman Sultans and their armies. The most obvious

position was one that European observers commented upon (and sometimes recorded in

illustrations), that of musicians leading the armies (a role that remains today in the modern Turkish

army for many Gypsies undertaking their military service).

The tradition of music-making and the training of children in the Gypsy community through the

guild system produced musicians and dancers of the highest quality that serviced the elites of the

palace and pashas' households throughout the Ottoman period. These dancers and musicians were

highly paid and extremely important in the daily life of the Sultan and his court. Gypsy women also

worked as story-tellers, seamstresses and washer-women for the palace harem, acting as mediators

with the external world for the secluded women of the palace schools. Coffee-house puppeteers and

purveyors of folk tales, the Gypsies occupied a specific and organised place in the complex ethnic

and social mosaic that was Ottoman society. Music and dancing were so synonymous with Gypsy

identity that even groups of Armenian, Greek and Jewish dancing boys and girls were known as

"Gypsy" by dint of their occupation, and we find the evidence of these performers in numerous

Ottoman miniatures that depict the festivals, processions and feast-days of the Ottoman calendar.

The masters and mistresses of their crafts were located in the Sulukule quarter were the system of

educating and training was carried out through a complex `apprenticeship' process that saw young

children progressing to accomplished and respected musicians and dancers, then with age to the

leaders and teachers of the next generation.

Not only were Gypsies in this neighbourhood the dancers (the illustration here shows a Gypsy cengi dancer from the 17 ̀h century; Abdülcelil celebi Levni, c. 1720 The Book of Festivals) and musicians but horse-dealing and carting also played a valuable part in the local economy. Porters and basket-

211


makers too lived in the area, close by the huge markets of the Grand Bazaar and the Egyptian

Bazaar, and in the reign of Ibrahim I the guild of basket-makers paid for the reconstruction of the

pavilion that now bears their name on the shores of the Golden Horn (1643). In other parts of the

Ottoman Empire we have records of Gypsies as water-engineers (in Salonika for example) and as

miners in Bosnia and soldiers in the Kosovo, whilst in the city itself, Gypsies were part of the

numerous street-traders that plied their wares in the neighbourhoods and on the highways. The

cultural heritage remains in the neighbourhood and the community with the music and dance

aspects very much alive despite the difficulties of the recent past. The historical heritage of

Sulukule is one that must remain as the oldest Romani settlement, and as the place where Gypsy

identity has been moulded and formed in the complex shifts of time and empire...

212


7.2 Gypsies in the late Ottoman Empire; the impact of nationalism, westernisation and

modernisation upon the cingene, 1789-1878

"Furthermore, let not any viziers... voivodes... sipahis... or other... groups interfere in the affairs of the Gypsies... as

well as inflicting any injustice or oppression upon them by demanding more than this Imperial decree specifies. "

Ottoman Firman, 1684

"[The Gypsy rabble should]... without further formality be beaten with rods and branded on the back and for evermore

be banished from the territories of the entire [Upper Rhine]... Circle... " Frankfurt am Main, 1711

Before briefly examining the position of the Romanlar or Gypsies in the late Ottoman Empire, it is

important to note that Ottoman social categorisation was primarily based upon religion rather than

notions of ethnicity as such (Braude & Lewis 1982: vii), although the complex interplay of what

Makdisi has called the "... a hierarchical system of subordination... " (2002: 768) included categories

reflecting notions of class and identity. In the context of a discussion about Gypsies, the

categorisation upon some notions of an "ethnicity" would seem to be identifiably part of the

location of those peoples of the Empire who increasingly became `peripheral', `backward' and

Oriental (Makdisi, 2002: 772). As the nineteenth century progressed, these notions became both

more nuanced and clearly fixed with the development of Ottoman Orientalism, through engagement

with Western European imperialism and absorption of the essentially scientific racist paradigm

(Gillard, 1984: 20). This was expressed through the tendency of the Ottoman elite to see Istanbul as

the dynamic, progressive and reforming centre of a modern Ottoman Turkish polity, whilst the

provinces remained the epitome of backwardness, unreformed, spatially, culturally and temporally

distant (Makdisi, 2002: 770-1). In this, the elements of confessional and ethnic identity were

primary boundary markers for the ruling elite, locating Ottoman quiddity and culture, as the

Ottoman Porte attempted to define the component parts of the Empire's diverse and polyglot

population. The servants of the sultan at the end of the nineteenth century saw themselves in a

radically different light to that of the aDyän at the end of the eighteenth; modernisers, reformers,

urbane and literate they perceived themselves to be the harbingers of a new Ottomanism

(Osmanlihk), one that was coeval with Europe in its history, but determinedly Islamic and modem in its outlook and the apogee of development in the empire. The centre of this world-view was Istanbul, with the hinterlands playing a similar role to that of the colonies for the imperialist

powers; these were the lands of the "savage" and irredeemable, the haunt of the nomad and

uncivilised "tribes". The development of what Usaama Makdisi has described as "Ottoman

Orientalism" follows a similar trajectory to that of Western Orientalism, the necessity of engaging

with an imaginary "other" in order to project an image of "self". The concept of "nesting

213


Orientalisms", suggesting that every society orientalises its "other" in the process of developing a

national, or dominant ethnic identity, is most appropriate here (Bakic-Hayden, 1995: 918); it is their

response to imperialism and the invasive, unequal relationship of the imperial powers in Ottoman

Turkey in economic, social and political spheres.

In identifying Gypsies in the late Ottoman period, the terminology of the Ottoman tax collectors

and government officials makes it difficult to always be certain about categorisation. For example,

Copt or Kipti was a religious term often used as synonymous with Gypsies, based upon the state's

acceptance of this population as "Egyptian" in some way (Bond 1917: 3), a position which reflects

other Ottoman notions of all adherents of the Orthodox faith as Rümi, or Greek. Other Ottoman

terminology included Cingene, (this term is considered derogatory amongst modem Turkish Gypsy

communities) a term derived from either the notion of Gypsies coming from further east than the

Turks (i. e. cin or China; see Marushiakova & Popov, 2001: 36) and a sense of peripateticism in the

word gene, meaning "again" in Turkish, or more simply from the Persian cingan meaning dancer,

entertainer. The designation of Arabaci from the use of horse-drawn wagons (Turkish at arabasi)

supplies us with an idea of transport, of being involved in both movement and moving as an

occupation and a "life-style". Some terms were specific to particular groups within geographical

locations; Poýa or Bo§a for example is still a term in current use in Eastern Anatolia (though

contested), whilst Mirtip or Mitrip is found historically in Hakkari, Mardin, Siirt and southern Van

(Arabic for `musician'). The Armenian Patriarch identified a group described as Kinchors (an

Armenian word for Gypsies; see Trotter 1881, quoted in Karpat, 1985: 125) in the vilayets of

Erzerum, Bitlis and Van (excluding Siirt). This term is not reflected elsewhere in Ottoman

terminology but appears again at the Berlin Conference of 1878 after the Ottoman defeat by

Imperial Russia, albeit in an "unofficially" submitted note to the Plenipotentiaries from the

Patriarchate (Karpat, 1985: 125, n. 3). The term relates to the late 18th century exhortations of the

Armenian clergy to marry Armenian Gypsies (Posha, a term that is also, like cingene, considered

pejorative; modern communiteis use the term Lom to describe themselves), thereby bringing them

into the orbit of the Armenian people proper, as additional members of the "nation" in the

discussions about numbers of Armenians. The Ottoman census officials of 1831 noted that amongst

the groups in the Anadolu eyalet were Tahtaci, Abdallar and Alevis (Karpat, 1985: 112). The

contested nature of such designations in modern ethnology notwithstanding (Benninghaus, 2002),

the Abdallar (stemming from the Turkish word abdal meaning wandering, begging dervish and

connected to the term aptal for a simple, foolish person) were considered to be Gypsies. Modem

214


ethnology suggests them to be an admixture of Afghan-Turkic nomads inhabiting Anatolia or a

group similar to the Yenische of Germany, Resande of Scandinavia or Romanichals of England

(Alford Andrews & Benninghaus, 2002). The claims of these groups to be recognised as Romani

peoples in their own right reflect the assertion by modern Abdallar that they represent a distinct

Alevi Gypsy "tribe". The connection between the Tahtaci and the Abdallar may be clearly seen in

their shared Alevism rather than any perceived shared ethnicity; however the problems of hxtero-

identification mean that sometimes a variety of differing tribal, religious and ethnic groups were

considered to be Gypsies by both the Ottoman state and the wider population (Karpat, 1985: 114).

The Tahtaci, Laz, Yörüks and some Kurds have been and continue to be on occasion, categorised in

a variety of differing ways according to changing social and political circumstances. It must also be

recognised that the majority of Gypsies in the Ottoman Empire after 1831 were Muslim, and so fell

into this category in much of the official documentation, rather than being enumerated separately.

Economic and military occupations were also indicative, and in some instances we find Gypsies

categorised as derbenci or voynuks, irregular border guards (Malcolm 1996: 114). The earliest

military description is of the Gypsy sancak, a military-administrative unit (and not a geographical

definition), in the 16`h century eyalet of Rumeli and described in the Law for the Supervision of the

Sancak Gypsies promulgated in 1541 under Kanune Süleyman. This unit was deployed by the

Ottomans along the borderlands between Bosnia and Kosovo. The discussion about the continued

existence of the Gypsy sancak is problematic, as it is difficult to establish in Ottoman records after

this period (see Mutafcieva & Dimitrov, 1968). The population count of the 1520's, taken in the

European province of the empire, the eyalet of Rumeli, provides us with information on numbers of

Gypsies, locations of Gypsy communities and their religious `beliefs' or identities (Malcolm 1998:

206) in the earliest imperial times.

We find also them amongst communities of miners in the Novo Brdo and Trepca regions of Kosovo

as muleteers, and amongst the armourers and falconers around Prizren (Malcolm 1998: 102-3). The

reports of papal emissaries, legates, foreign travellers, ambassadors, Catholic and Orthodox clerics

and the occasional military dispatches from invading Habsburg commanders can also give

indications of the position and extent of the Gypsy populations (Malcolm 1996: 116), estimated to

be 8,000 in 1808 and 11,500 in the 1860's. Condemnation from Muslim state and religious leaders

(Hasluck 1948: 1-12; Soulis 1961: 155), or members of the ulema (Ginio, 2004: 117-144) also tell us

something about the ways in which Gypsies were perceived from an official perspective and give

some indication of changing notions. Their "traditional" skills as metal-workers and blacksmiths

215


means that Ottoman Gypsies can also be found working for the Sultan (miri) in various of the state

enterprises other than mining. The presence of Gypsy blacksmiths in the Arsenal in Kasimpasa may

date to the establishment of the naval manufactory, but certainly the increasing use of "itinerant"

blacksmiths and overall reduction in the numbers of those who could claim a "fixed" position (miri

ahengeran) in the eighteenth century meant a rise in the numbers of Kiptiyans in the employ of the

state, albeit temporarily (Faroqhi, 2001: 159). Hired when the advent of a campaign was imminent,

the rest of the time these smiths earned a precarious living in neighbourhoods like Kasimpasa,

Kaghithane, Tophane and Balat by a variety of means. The notion that Ottoman Gypsies can be

found in very particular occupational `niches' however, reflects the inappropriate application of

experiences documented elsewhere upon the Ottoman society. Many of the blacksmiths working in

the Arsenal were not Kipti or cingene but shared a social and economic position just a precarious. A

petition from these men and their wives in 1720, demonstrates that poverty amongst this group was

such an issue that they appealed to the Sultan directly, this being the best means of resolving a

dispute with another group (Faroqhi, 2001: 157). The blacksmith's wives engaged in the cleaning

and selling of lamb's trotters in competition with the male guild of pacaci taifesi to supplement their

meagre incomes and feed their children from the proceeds, as it states in the document submitted to

the Padishah (Faroqhi, 2001: 155). Whether the Kiptiyan women also engaged in this practice is not

recorded, although the Ottoman practice of listing Gypsies separately would suggest that they were

not.

One can find Kiptiyan women recorded as performers at weddings; as singers and players they were

present in Ottoman Cairo in nineteenth-century Egypt and in other major cities too. Istanbul's

dancing guilds had become so synonymous with Gypsies that all dancers, whether Romani or not,

were described as cingene. Competition between groups of Jewish, Armenian and Greek cengi dancers sometimes sparked open conflict between them, to the despair of the authorities. The

Ottoman population records for the 19th century provide much better data for Gypsies, despite the

prejudices (both then and now) of those demographers determined to suggest them inadequate

(Karpat, 1985: 108-224).

In this section I shall present two `snapshots' to illustrate some changing trends in the general

situation of the Gypsies under Ottomans rule during the 19th century. I will show that their treatment in the empire, whilst comparatively better for Gypsies throughout most of its existence,

significantly worsened by the end of the century with increasing European influence, and this had important consequences.

216


During the period from the end of the 18th century to the end of the 19th century the Ottoman

Empire was undergoing continuing processes of internal reform and economic dislocation,

concomitant with increasing Westernisation and modernisation (Davison 1990: xi-xv; 73-95). This

can be seen as both Ottoman initiative and a response to European pressure and increasing

penetration of Ottoman markets by France, Britain and other Western states (Quataert 1993: 6-7).

This included the efforts by Europeans to secure effective control of the Ottoman economy through

banking, especially after the financial collapse of 1878. Territorially the empire was becoming more

compact as a result of military defeats at the hands of the Imperial Russian armies (Davison

1990: 21), whilst its own officials removed provinces and regions from effective control of the

Dergäh-i all or Sublime Porte, sometimes permanently (for example, Muhammed Ali and Egypt).

The subject populations making up the empire's complex ethnic mosaic also began to feel the

impact of the new ideologies of nationalism, self-identification and ethnic self-ascription, often

through European intervention as in the case of Russia and the Orthodox populations (Davison

1990: 29-43) or via the growing mercantile classes which were predominantly, though by no means

exclusively, non-Muslim (see Stoianovich 1960). This manifested itself in movements breaking

away to found national states from 1804. Increasing contact with Europe from the accession of

Sultans Selim III (1789-1807) and Mahmüd II (1808-1839), and the demands of the tanzimat

modernisation programme (Tanzimat-i Hayriye or `beneficial re-ordering'), after the promulgation

of the Gülhane Hatt-i Sherif (or Imperial Rescript of the Rose Garden) in 1839, meant larger

numbers of foreign engineers, military officers and diplomats came to be resident in the empire than

previously had been the case. Ottoman diplomatic and technical overtures to European powers

meant that borders, both geographical and ideological, between the dar-ül Islam or the house of

Islam, and the dar-ül harb or the house of war, became more permeable than ever before (Lewis

1982: 51-57).

There are a number of key groups in the Ottoman social structure at the end of the eighteenth and

beginning of the 19`h century. The traditional military forces of the empire, the yeni Geri or

janissaries (celebi 1951) had become a largely ineffective militia with integral links to the guilds

and artisan class. The janissaries maintained power through alliance with the conservative elements

of the 'ulemä, to oppose reform and the centralising tendencies of the government at the opening of

the century (Shaw 1971: 367-383); although the 'ulemä could also be a force for Westernisation and

modernisation at certain times, especially through the Mevlevi orders (Heyd 1961: 63-65). The

Christian churches, with their sponsorship by European powers especially Russia, had achieved a

217


degree of prominence previously unknown (Davison 1976: 465). The traditional divide between the

mass of the population, the reäyä or both Muslim and Christian peasant-producers, and the military-

bureaucratic elite or umerd, had broken down to a significant degree in the 17th century,

undermining the ideological and fiscal basis of the state (Kunt 1983: 67). Most essentially in the

context of a discussion about Gypsies, the growth of an important group of intermediaries called the

a'yän or `notables' had taken place in the course of the 17`h and 18'h centuries (Barkey 1994: 1-8;

McGowan 1994: 637-758).

The aDyän combined two distinct elements - the urban, tax-farming administrative notables,

essentially members of the old umerä class and the warlords controlling vast numbers of irregular

troops which they often, though not always supplied to the imperial government in times of war

(Inalcik 1977: 27-52; Karpat 1972: 243). These had fused into one class, combining the functions of

both warlord, tax-farmer and government official by the closing decades of the century and, in

emulation of the Sultän's household became microcosms of the court at Istanbul, occasionally to the

extent of conducting foreign negotiations independently of the Porte, as in the case of 0A171

Pasühfla Tepedelenli of Yanya (Ioannina) (see Bowen, 1986: 398b; Davenport, 1837: 87-90). Their

economic power lay in their ability to ameliorate the tax demands of the treasury or even shelter

whole communities from the state fisc in return for extracting significant surpluses (McGowan

1994: 659). Able to enforce their will both locally and regionally and constantly involved in attempts

to extend both influence and power, the aDyän were almost always resistent to attempts at reform or

centralisation by government which meant a reduction in their own authority and on at least one

occasion in 1808, were able to dictate terms to the Ottoman state itself (Shaw 1971: 404-405). The

development of power relations between this group and the state is characterised by de-

centralisation of sultdnic authority, sometimes to an almost fatal degree under Sultan Selim III and a

re-feudalisation of provincial power and social relations (McGowan 1994: 658). However, the state less often met this resistance with outright force, using bargaining, offers of pardons and even

government appointments (Barkey 1994: 2-9) in attempting to deal with the aUyän. This group in

the late 18th and early 19th centuries, had a particular impact upon the Gypsies of Ottoman Europe.

The Gypsy communities in Balkan towns and cities occupied particular mahalla or neighbourhoods, in common with religious groups such as Orthodox, Catholics and Jews and guilds of tanners,

weavers and carpet-makers and others (see Journal of the Gypsy Lore Society Old Series, 1: 125,

3: 29; New Series, 1: 198,7: 158,8: 158; Third Series, 38: 154). In some sandjaks or provinces like

Macedonia, parts of Bulgaria and Kosovo, there are records of entire Gypsy villages (Paspates

218


1870: 11; Bond 1917: 5). In militarily strategic garrison towns such as Belgrade, Gypsy communities

were situated around the fortress, although in many towns they had been pushed to the edges of

settlements. In matters of state revenues, it is not clear whether the Gypsy sandjak created in the

mid-16th century for purposes of taxation, was still in operation at this time (Malcolm 1998: 206). In

general Gypsies both Muslim and Christian, tended to be taxed at a higher rate than the reaya

(literally "flock" - an indication of the Ottomans' own nomadic origins) or peasantry, seemingly on

the basis of religious adherence or lack of it (see below), but possibly for sound economic reasons

(Soulis 1959: 154-156). The state's suspicion of nomads in general (see Lindner 1983: 7-14; Inalcik,

1996: 37-41) and the difficulties of extracting taxes from originally mobile or semi-mobile

populations may also have contributed to an annual collection ensured to raise maximum revenues

often in kind or nüzul bedeli (Faroqhi 1994: 533-535). Tax rates for the resm-i sift or land tax (paid

by Christians as harac, or ispence in parts of Serbia and Kosovo), in the earlier Ottoman period

were 22 akces for Muslims and 25 akces for Christians, with an occasional additional 6 akces,

payable as the head-tax or cizye on the latter. There seems to have been an increase in these rates at

the end of the 17`h century and beginning of the 18th probably due to the increasing cash shortages

experienced by the Ottoman treasury (Soulis 1959: 155). However, by 1750 CE the cadastral

registers (defters) had become generally inaccurate, forcing a change to a district `re-allocation' or

tevzi system by the close of the 18t" century (although the change was organic and uneven in

development; see Inalcik, 1996: 55-72). This allowed the imposition of any new fiscal demands to

be ameliorated within the district based on realistic assessments of productive capability by the

local all yän, thereby strengthening their control (McGowan 1994: 714). Ottoman fiscal policy

continued to tax the countryside in favour of the town and revenue sources remained largely

consistent and conservative (Quataert 1994: 870). Horses and other livestock, for example went

untaxed in some Rumelian provinces, a factor benefiting Gypsies and Vlachs, whilst potential new

sources were not exploited. As smiths, metal-workers, gun-makers and repairers, armourers,

powder-makers and miners, Gypsies had a high economic value placed upon them and these

occupations were granted certain tax-exemptions by Ottoman officials often amounting to relief from the avariz or extraordinary taxes paid in cash, annually collected by the mid-17`x' century (Faroqhi 1994: 35).

Those Gypsies involved in military occupations, like other derbenci and yürüks, also enjoyed

privileges, as in the case of the Gypsies who defended Kosovo against the Habsburg armies in 1737

or Bosnia in 1788 (Malcolm 1998: 207). These Gypsies may have been largely sedentarised as

219


recruitment from nomadic groups would have been difficult; although seasonal campaigning was

the norm for nomadic and semi-nomadic peoples who have generally combined both a military

function and mobile lifestyle very successfully (the Avars, Huns, Mongols and of course Ottomans

are examples). The Ottoman organisation of Gypsies and Yürtiks into hereditary ocaks (groups of

irregular troops about thirty in number), is recorded by both Evliya celebi and Sir Paul Rycaut in

the 17th century (Halliday 1922: 170-171). It is in this context that numbers of Gypsies were

attached to the a0yän (at least some of whom may have been Gypsy in origin themselves, as Ali

Tepedeleni was rumoured to be). QAli PasOhüa Tepedelenli of Ioannina, QOtDhDmän Paswan-

ogOhElu of Vidin (1758-1807) and Ismail Aga of Rus6uk all had significant Gypsy communities

attached to their retinues (Halliday 1922: 168; Bowen 1986: 398b). Metin-Kunt's argument that the

dev*irme was neither restricted to Christian's nor to the Porte, but was part of the relationship of

patronage that made up Ottoman social organisation, may also have a bearing here as the ethnic-

regional origins of Ottoman officials was often reflected in these official's personal retinues (Metin-

Kunt 1982: 60-63). QAli PasOhDa's campaigns against the BuDath dynasty of $kodra record the

presence of Gypsies, Yürüks and Vlachs amongst his forces (Bowen 1986: 398b). In the re-

feudalised order of the late 18th century Balkans, Gypsies played an important role in craft

production, horse-trading, mining and metallurgical economies and militarily for both state and

aD yän.

As well as these groups, some Gypsies were in the position of being slaves, though not in the same

sense as those of the Romanian principalities (Georghe 1983: 13-22, Fraser 1992: 131-145, Acton

1993: 77-89). A small number of Gypsy miners and metal-workers are present as ̀ slaves' in the tax

defters, though it is not clear what the circumstances of their enslavement were (Malcolm

1998: 207). It is possible that these were the remnants of enslaved communities captured by the

invading Ottoman forces in the 16th century, such as the mining community at Novo Brdo, but it is

also the case that sections of this community were encouraged to continue mining through tax

exemption privileges, though likely that they had no freedom to emigrate (Malcolm 1998: 102).

There is no record, at present, of the kind of serfdom comparable with Moldavia and Wallachia for

Gypsies in the Ottoman Empire as a whole, but as European commentators noted in the mid-19th

century about the community of Gypsies that inhabited Ali Tepedeleni's ruined konak or mansion at loanina, their condition could be equally one of extreme poverty and destitution (Hughes 1830,

1: 455).

Nomadic and semi-nomadic Gypsies in Rumeli or Ottoman Europe were occupied as horse-traders,

220


itinerant metal-workers and tinners, basket-makers, spoon-makers, comb-makers, dancers and

musicians (Petulengro 1915: 6-28). Some appear as "necromancers" in Catholic prelates' reports,

practising the arts of fortune-tellers, healers, cunning-men and women, and folk-doctors (Malcolm

1998: 130). In common with the general rural picture, shared beliefs and practices encompassed a

genuine syncretism between Islam and Christianity for the mass of the population, whilst the

continuing pagan practices of pre-Christian and pre-Islamic beliefs were incorporated into the folk-

religion predominant in the Balkans region (Malcolm 1998: 129-134). Gypsies were part of this

syncretism of beliefs and the idea of them representing `heretical' Muslims or not being `good'

Christians was a common conception amongst the urban, educated priests and mollahs, and true of

the rural population in general. Another shared attribute is the `Islamisation' of some Gypsy men,

whilst women remained Christian. Whether this led to the phenomenon of `crypto-Christianity'

amongst Gypsy communities, as it did amongst other rural communities is not known. However, as

well as the Christian clergy, the Muslim 'ulemä of Salonika had cause to condemn the Gypsies of

the surrounding area for their lack of faith and indifference to religious observance (Halliday

1922: 177). Widespread celebration of both Christian and Muslim feast-days involved Gypsies as

dancers and musicians (during the Feast of St. George and the Assumption of the Virgin), with

important consequences for their position as bearers of `popular' culture in post-Ottoman Serbia

(discussed below). The question of religion amongst Balkan Gypsy communities in the late

eighteenth and early nineteenth centuries must be viewed in the context of the general picture of

syncretism amongst the rural population at large (Malcolm 1998: 129-134).

Two important factors emerge from this brief overview; namely the relationship between the aUyän

and Gypsies, and the sedentary character of many of their communities. The numbers of nomadic

and semi-nomadic Gypsies in the Balkans at this time would appear to be smaller than later in the

century. Gypsies look to be more closely integrated into Ottoman Balkan society and economy at

this point than after the middle years of the nineteenth century, and less mobile. This picture stands

in contrast to accounts of travellers and "Gypsiologists" of the 1880's onwards, where emphasis is

placed on the view of Gypsies as travellers and nomads. The reasons for this may lie in the personal

bias of those investigating Gypsies in later years who, under the influence of an idealised

romanticism, sought to find a ̀ traditional' society of wandering people, largely unsullied by

industrialism and `modem' society. In the context of the `Orientalism' prevalent in European states

during this period (Said 1978; cirakman, 2002: 1-32), this image of the Gypsies falls neatly into

these perceptions. Therefore, the seeking of non-nomadic Gypsy communities, established for long

221


periods of time fails to fulfil the aspirations and stereotypes of those seeking the `traditional' and

`genuine' Gypsy. It is also the case that Gypsy communities of long standing in some areas became

acculturated, in common with other groups, losing language and customs in favour of the dominant

culture, and so less easy to identify (Malcolm 1998: 208).

The empire at the close of the 18`h and beginning of the 19`h centuries was one in which the endemic

banditry, conflict between state and aflyän and the diminished central authority of the Sultän and

his government, left existence in the provinces difficult not to say precarious (Barkey, 1994: 2).

Travellers' tales of highway robbers and bandits attest to the frequency with which travel was

endangered by such depredations (Hughes 1830: 2-352). Constant conflict between competing

aElyän and their retinues also contributed to what could be argued was a state of degenerative

anarchy at this time (Shaw 1971: 211-246). With war throughout the century, the process became

accelerated as warlords sought to increase their influence at the expense of their neighbours, assured

of the inability of central government to intervene (McGowan 1994: 663). These conditions made

nomadism hazardous and difficult except within very localized areas, usually under the effective

control of a provincial warlord, or Ottoman pa§a (areas under the control of local allyän were

surprisingly peaceful; Ali Tepedeleni mercilessly crushed banditry in his area and actively

encouraged trade); Shaw 1971: 228-230). The value of Gypsy smiths and metal-workers to the

aDyän was one which made it important for them to try to discourage movement and loss of these

valuable skills. These groups of Gypsies may have been more tightly controlled by those aDyän for

whom they performed these services, as seems to be the case with the Gypsy community at

Ioannina (Hughes 1830,1: 455).

A general suspicion of nomads was part of the character of the Ottoman state, as suggested above.

Able to cross borders, difficult to control during and after campaigns and to tax, often unsuitable as

enforced settlers on abandoned lands (under Murad IV for example; Roux 1987: 229), nomads could

be perceived as smugglers, gun-runners, illegal traders with the empire's enemies and sources of

social discontent, heretical belief and crime. Gypsies, Vlachs, Türkmens and Tartars all attracted the

attention of Sultans, papas and viziers who formulated and promulgated legislation designed to

curtail or limit the movements of whole sections of the population (including at times, women), in a

variety of ways (Fraser 1992: 176). Finally, the economic dislocation of the Ottoman market by

European powers had not reached the proportions it would during the later century, causing

widespread social dislocation and economic migrancy. The loss of imperial territory, whilst of grave

concern for the state, had not had the major impact it would when western-sponsored Balkan

222


successor states came into being, causing large-scale population movement and demographic

disorder (Quataert 1994: 777-795; Karpat 1985: 156). These factors would seem to indicate reasons

why the incidence of Gypsy nomadism was less frequent than later in the century.

The mid to late 19th century saw a drastic alteration in this picture. Power relations between the

a'ydn and the Porte had been largely reversed in favour of the state by means of a major and largely

successful centralisation policy by Mahmüd II (1808-1839). The destruction and subjugation of

alternative sources of authority and those resistant to reform was the leitmotif of Mahmüd's reign,

and in all but the case of his Egyptian vassal, Mohammed All, the Kurdish ämirs or princes and

parts of the Arab lands, he largely achieved these goals (Lewis 1961: 76-106). From 1813 onwards,

Mahmüd and his competent viziers struck a series of devastating blows to those ayän who had so

successfully resisted his predecessor, Selim III. By the time Mahmüd destroyed the janissaries in

1826 at Et Meydan in Istanbul, he had eradicated major a'ydn such as All Tepedeleni and Ishmail

Patzvanoglu, or encouraged them to identify their best interests as laying with the state, as with the

Karaosmanoglu beys of Anatolia (Zürcher 1994: 32-51). The period of reform in the empire known

as the "tanzimat" dates from the end of Mahmüd's reign. The promulgation of European-influenced

legislation in the 1839 Gülhane Hatt-i Sherif granted equality before the law for all Ottoman

subjects (although it was not clearly formulated on this point) and was further strengthened by the

1856 Hatt-i Humayun (Zürcher 1994: 52-74). This was also the point at which the reformers found it

prudent to prohibit slavery in the empire (although it was never formally abolished), under intense

British pressure (Erdem 1996: 67-124), whilst Gypsy serfdom was abolished in the Rumanian

principalities in the same period (Georghe 1983: 12). This era of reform and transformation

continued until 1878 and the end of the First Constitutional government (1876-1878). This is a

period (1826-1870) of massive economic penetration by European powers, in particular by British

and French capital, when the Ottoman economy becomes "semi-colonial", a producer of raw

materials and buyer of manufactured goods. The dislocation and re-structuring of Ottoman society

gave rise to the increasing division between a largely Muslim military and bureaucratic elite and an

economically powerful Christian bourgeoisie (Quataert 1994: 889-894).

The growing impact of nationalism and nationalist ideologies is of central importance to the empire

and its subject peoples. The Serbian revolts of 1804 and 1813 led to autonomy for the Serbs in 1815

and independence in 1878. The Greek independence movement and war with the Ottomans from

1821-1827 resulted in the establishment of a small Greek state that was able to expand its territory

through war in 1864 and again in 1881. The unification of Wallachia and Moldavia in 1859 formed

223


the basis for the independent Rumanian state of 1878, whilst this date also saw Bulgaria achieve the

status of a principality and Bosnia jointly administered by the Austro-Hungarian Empire. The

Russo-Turkish War of 1877-1878, and the settlement that followed (the Treaty of Berlin), lay the

foundation for a fundamental re-drawing of the Balkan map, re-aligning European power vis-ä-vis

the Empire. The formation of the successor states resulted in huge territorial loss and an ethnic

homogenisation, as Muslims were expelled from the new national states and Christian populations

moved into Serbia, Bulgaria and Rumania (Jelavich 1983: 2; 254-62).

The question of agricultural dispossession, once thought to be an important element of `big-estate"

or ciftlik formation, is central to the question of Gypsy mobility. Dispossession in the latter half of

the 19th century most frequently involved nomadic and semi-nomadic peoples with customary rights

to the land they used. Competition between sedentary communities and these groups was

exacerbated by the state appropriation of these irregularly cultivated areas and pasturage, seen as

waste or mevat land in economic conditions of profitable development, driven by foreign demand

for Ottoman agricultural products (Quataert 1994: 873-874). Once the process of `reclaiming' this

land had begun in the more marginal but recently pacified regions of the empire, it soon spread to

the central lands. The net result was increasing dispossession of land by groups of peasant and bourgeois entrepreneurs at the expense of Kurds, Türkmens, Tartars and Gypsies.

The rejection of non-national identities and pressure to conform to national ideologies or become

assimilated into the new national communities was of great import to the Balkan Gypsy population. The important role of nationalist myth-making involved (and still involves) a degree of re-discovery

or `re-invention' of traditions and culture identified as pertaining to `the people' (Hobsbawm

1983: 76-83). In this context, Gypsies were associated with Balkan popular culture as entertainers, dancers and musicians. In Serbia in the 1850's for example, Gypsies were seen to be a part of this

process (Malcolm 1998: 208). At the same time, other less positive elements of the national identity

were also being `forged' - Gypsies as thieves, vagabonds, betrayers of the Serbs associated with Vuk Brankovich at Kosovo in 1389 (Pennington & Levi 1984: 14-17) and àlien', òther', non- Serb/Bulgarian/Greek. A similar process also took place in relation to Ottoman Jews (Dumont 1982: 209-242). Nationalist literature assigned (and continues to assign) these other groups specific roles; Gypsies became the hangmen of those `liberators' and `patriots' who have struggled against the Òttoman yoke' and suffered a martyr's death; or they were spies and traitors working for the "Turks" (Andric 1994). Western European linguistic analysis of Romani languages, especially in Germany and Britain fueled the identification of Gypsies as part of the òther' and Òriental' (Okely

224


1983: 8-13), in contradistinction with those ̀ true' Europeans who have either defended the continent

against the `Asiatic hordes' (as in the Croatian ideological trope of antemurale Christianitatis), or

those who sacrificed blood and nationhood to save Europe from alien domination (most clearly

expressed in the Serbian mythology constructed around the 2°d Battle of Kosovo in 1389; Malcolm

1994: 20; 1998: 58-80). This dual and contradictory identity - as holders of elements of folk-culture

and as alien - has had repercussions lasting to the present.

This period also saw increasing intolerance towards non-Muslim subjects of the empire and the

construction of Osmanlilik, the notion of a Muslim Osmanli identity, in competition with pan-

Turanian or Turkicist nationalisms. The apogee of the former was reached under Abdülhamid II,

who asserted claims to the Caliphate as an antidote to nationalism and modern secularism prevalent

in the empire (Zürcher 1994: 83). This development had an impact upon Gypsies, Jews and

Armenians, increasing antagonism towards non-Muslim Ottomans, with appalling consequences for

the Armenian community in the early 20t' century. Whilst suffering a less intensive oppression, both

Gypsies and Jews, especially as peddlers and itinerant workers, experienced much popular violence

against them as perceived carriers of epidemics, child-stealers, thieves and vagabonds (Dumont

1983: 223).

In the later 19th century, a marked deterioration in the position of Ottoman Gypsies can be seen. The

destruction of the a'yän and their formidable retinues forced those Gypsies attached to them to re-

establish nomadic and semi-nomadic patterns. The increasing economic dislocation resulting from

European penetration of the Ottoman markets had a proportionately greater effect upon those

communities, like the Gypsies, who had a particular economic position within the predominantly

agrarian Ottoman economy. They became marginalised, further contributing to larger numbers of

them becoming migratory in order to exploit what small `niches' remained in a rapidly changing

economic and industrial structure of Ottoman Europe; though there is some question about the

predominance of economic factors in Anatolia (Roux 1987: 229). Agrarian dispossession inevitably

exacerbated these trends. Nationalism and the formation of the new nation-states, with their ideology of ethnic homogeneity and expulsion of `alien' populations, had a strong impact upon Gypsies forced to adopt contradictory stereotypes and identities or be assimilated. Nomadism, in

this case, was a means of resistance and survival. Increasing intolerance from Ottoman society towards Gypsies and other groups added to the pressures already extant upon Gypsy communities in the Balkans and Anatolia.

The question of migration in this period is an important one; Europe experienced a second major

225


influx of Gypsies in something that appears, initially at least to resemble the earlier migrations of

the late 14th and early 15'h centuries and to be related to the release of large numbers of Gypsies

from slavery in the Rumanian lands. Recent scholarship has argued against this factor as the driving

force behind Gypsy migrations of the late 19th century and suggests that the cause may lie

somewhere else. In fact, large-scale migration due primarily to economic factors was a

characteristic of late Ottoman society in general (Karpat 1985: 15). The empire's population,

particularly its non-Muslims, left the territories of the Porte in such large numbers that at times it

resembled an exodus (Karpat 1983: 387). Gypsies, Jews, Armenians, Georgians, Syrian Christians

and Anatolian Greeks, made their way to those parts of the world with more economic opportunities

and an improved security of life than could be found in late Ottoman society. In this context, Gypsy

emigration was clearly part of the wider social and economic conditions experienced by the

Ottoman Empire and Mediterranean world at this time (Quataert 1994: 791).

As the empire became more culturally and ethnically homogenous and territoriallycompact, Gypsies

became increasingly nomadic as a means of exploiting the limited opportunities that existed at the

economic edges of Ottoman society. This more marginalised position was also a reflected in Balkan

societies in general, together with increasingly antagonistic and contradictory stereotyping and

widespread discrimination. Notions of the Gypsies as clearly `alien' in the way that the Jews also

experienced were commonplace. Identified with being carriers of disease was part of the

increasingly `biological' language of scientific racism, most often developed and deployed in

support of the new nationalisms. As these images became part of the `national consciousness' of

ethnic identities in the emerging national states of the region, the economic and social position of

Gypsies worsened. These ideological constructs lay behind the murder of many thousands of

Gypsies by the Ustasha regime of Croatia and Serb extremists in Bosnia (where Gypsies were

frequently identified as Muslims or Turks) during the 1941-45 period. Ultimately, it is these trends

in the late Ottoman Empire and its successor states that are now playing out their legacy in

Rumania, Bulgaria, Albania, Serbia, Bosnia, Kosovo and south eastern Europe today.

226


8.1 Time, narratives of journey and the writing of Gypsy history

Nasreddin Hoca came to a crossroads where two carts had collided. The drivers were arguing

furiously over whose fault the accident had been, and appealed to the hoca [teacher] to resolve the

issue. He dismounted from his donkey and sat listening first to one man, who told him of the event,

and then the other. He thought for a while and then said that both were right, when his donkey (a

clever beast) looked up and remarked, "But hocam, both of these men cannot be right. " The wise

old man smiled and said, "And you are right too... " (Sheikh Ahmet Karacadag, Diyabakir April

2007)

THE QUESTIONS IN THIS BOOK THAT I HAVE ATTEMPTED TO ADDRESS REFLECT, in some ways the story

about Nasreddin Hoca, whose tales and folk-wisdom are still widely appreciated in Turkey. The

primary question has been one of authenticity or rather the notions surrounding the production of

narratives about history, ethnicity and identity that implicitly or explicitly claim certain kinds of

authenticity and authority. These can be based in notions of scientific objectivity, through various

devices such as the use of the `third voice' and the distancing of the author from the text through the

use of the impersonal. The appeal to science as an objective and impartial representation of the facts

is one that has implications not only in the writing of Romani history, but also in a wider context of

access to resources, in that these should be awarded on the basis of such notions as ̀ science' being

carried out by `scientists'. In contradistinction are those whose scholarship is `tainted' through

engagement with activism and a political agenda, the so-called "web-page polemics" referred to

earlier in this book. The authenticity claimed by these writers is one based in the notion of the

genuine ̀ ethnic' voice, the indisputable perspective of experience from `the inside'. These positions

are seen as oppositional to each other, a choice of either/or where both `parties' will draw upon

fragments of the past (usually linguistic, or documentary) and construct a narrative that then is

argued over as if it represents an actual, tangible reality. The propositions here are that both or all of

these competing versions are constructed and reconstructed through successive reinterpretations,

that the past itself is irrecoverable precisely because it is the past and that all attempts to portray it

in the present are constructions, narratives that take these fragments and literally weave together a

seemingly coherent version of events, using particular tropes and emplotments, that then allow us to

justify the conclusions we draw as historians and scholars. This is not, of course restricted to the

writing of Gypsy history or histories, all history or histories are constructed narratives and reflect

the concerns, ideas and notions of the present to a far greater degree than we would perhaps wish to

recognise. Even the pursuit of the past, the areas of enquiry that we choose are related to these

227


concerns and the historiography of history writing is one that illustrates the shifts and changes that

have taken place in these concerns over time (see Pitterberg 2003; Robinson, 2003; Ersanh, 2002;

Lucassen, 1993). This seems to leave us in the situation of having dismantled the entire edifice of

historical scholarship and leave us without any structure, or even negate the proposition of

scholarship itself (Domanska, 1998: 173-81). However, it is in the very recognition of the

arbitrariness of the endeavour that the solution lies, as to admit that the exercise is flawed is not to

say that it is pointless. The justification for enquiry is no longer to be found in the grand narratives

of imperialism, colonialism and modernism, or the deconstructive assault on these of post-

modernism, but in the much older principles that suggest the pursuit of knowledge is about

understanding and informing ourselves and others, making sense of the world that we live in. The

critique about the kind of knowledge that is pursued, the ways enquiries are undertaken and the uses

of that knowledge and how it is presented allow us to see more clearly but do not, I would argue

undermine the purpose itself.

What of the understanding of Romani Studies? The current questions surrounding ethnicity, identity

and history are reflective of the wider notions that exist, the seeking for definitions of group and

belonging that have been at the heart of the human experience since the Athenians first sought to

define the barbarians', and articulate notions of who and what they were (Hall, 1989; Ascherson,

1996). The particular circumstances might be seen in the context of resources to support research

where questions of ethnicity and identity are bound to modern notions of citizenship and social

inclusion, and more specifically to the appropriation of the past by meta-narratives of Western

modernity and eastern ̀ backwardness', in pursuit of a unitary history that ultimately shapes the

future (Sardar, 1992: 496). The period when resources were being provided to support more activist

programmes have largely come to an end and those who have been part of the mobilisation of

ethnicity amongst the Gypsy peoples have now to make a shift that brings them, in some ways into

competition with the research and academic communities. Herein lies the basis for the arguments

that wash back and forth between the two sectors - the claims to authenticity and authority being

mobilised in search of more `legitimate' grounds to secure resources. The actual dynamics are of

course, much more complex as the relationship between scholarship, research and activism in

Romani Studies (as in many other areas) is interdependent and the needs of activism frequently acts

as a spur to scholarship, as Matras has illustrated (2004: 68-9). Scholarship and research provides

the basis for much of the impetus behind activism by continuing to refine and define the

components of Gypsy identities, ethnicities and histories. It is a symbiotic relationship where the

228


apparent conflicts are themselves necessary stimulus to further research and activism and this book

is a product in many ways of these conflicts.

So, if all history is a constructed narrative built around fragments of the past that are consciously or

unconsciously chosen in relation to the concerns of the present (whichever present that is), and then

emplotted using tropes that conform to wider understandings of the circumstances in which the text

is being produced, this book is clearly part of these processes, as I have tried to demonstrate. It is an

attempt to expand upon the notion of complex Hindia origins and reconcile the apparent differences

between the seemingly oppositional positions of the kshatriya warrior-theory and the notion of a

pariah or outcast origin, by claiming that both are compatible with the version of the usable past I

present. I have also tried to look more closely at the questions surrounding these notions of origins,

by adopting a more usual historical approach in examining the context for the ethnogenesis of the

Gypsy peoples. This has meant that I have presented a range of sometimes speculative

considerations, based upon the documentary sources (chosen to support such speculations of course,

in my overall emplotment), to argue for a more plausible (making clear the ambiguity of this term in

its original meanings of both `believable' and `applauded') account of Gypsy history in its early

stages, that might account for the various stages of movement and migration and provide possible

reasons for these taking place. I have also tried to present a series of considerations that reflect my

underlying `agenda', namely that the Gypsies of the Turkish and Arab lands are historically some of

the oldest communities of Gypsies in the world, yet scholarship has largely neglected them until

recently. The focus has consistently been on those Roma, Dom and Lom in the quondam south-east European successor states to the Ottoman Empire: Bulgaria, Romania, Hungary, Serbia, Bosnia,

Albania, Kosovo, Montenegro, Macedonia and Greece. This "fault line" has been almost

impermeable, and the Near and Middle East are even less researched or discussed in the

scholarship. The Gypsies of these regions merit almost no attention in anything other than specialist journals (Kuri: the Journal of the Dom Research Center), isolated articles and monographs (Streck,

2004: 179-94; Hanna, 1983,1982), or the press and media of the individual countries in which they live. The material in the latter is frequently uninformed and often deeply prejudicial. Despite the

relatively large numbers of Gypsies, as a percentage of population in the former Arab lands of the Ottoman Empire, their location, languages, culture and religious practices would appear to have

been of almost no interest to post-war Romani Studies scholarship. Outside of that produced by

evangelical organisations (although not all evangelise amongst the Gypsies), and some local

government statistics (Kagithane Belediye in Istanbul produces useful information about the

229


Gypsies in the municipality), Romani Studies scholarship has delivered relatively few reliable

sources of information for these groups until recently (the ERRC/hCa/EDROM research project,

"Promoting Roma Rights in Turkey" has provided the most up-to-date survey; see Marsh, 2008a).

The situation in the Caucasus and central Asia is almost equally bleak; general information about

the Lom and other Gypsy groups from any of the territories of these turbulent regions is reduced to

scant notes (Thubron, 1994: 92) and photographic volumes (Kuznetsova, 1998). The "Other"

Gypsies may feature in the work of linguists, anthropologists and ethnologists (for example, Akiner,

2004), but not many Romani Studies scholars. The obvious question is why has this been the case?

The traditional perspective of Romani history has been poised about two points. The early 20`h

century work of John Sampson identified a single "exodus" of the Romany peoples, which, he

argued subsequently subdivided somewhere in the region that is now eastern Persia (Sampson,

1923: 156-69). The linguistic evidence for this was the basis of Sampson's argument, in which he

attempted top demonstrate a clear relationship between Romani, Domart and Lomavren. However,

this was challenged soon after, by Sir Ralph Turner (1926: 145-89), a prominent linguist and

"Gypsiologist", he was also director of the School for Oriental & African Studies during the 1930's.

The relationship between Romani, Doman and Lomavren was not as close as Sampson had

suggested he argued, severely undermining this "single origin" (monogenesis) book. Turner's

speculations about the differences between Romani and Domari were more carefully argued than

merely to question the relationship between the two; he constructed a complex account explaining

the emergence of Romani as an Indo-Aryan tongue that had originally migrated from central India

into north-western Multan and Rajasthan before 250BC, on the basis of certain lexicological and

phonological features preserved in Romani but altered in related dialects in the `homelands' of the

central regions (Fraser, 1992: 21). Turner still opted for a relatively early departure of the proto-

Romani speakers prior to the 9th century AD, much as Samson had done. The major difference lies

in the question of origins; Samson maintained his insistence that Romani was a product of the

developments in north-western India in a reply to Turner's critique (Sampson, 1927: 57-68).

Turner's own response continued the theme of a central origin for proto-Romani (Turner, 1927: 129-

138). Other scholars interpreted the evidence differently positing wildly disparate dates of departure

(Kaufman, 1984; Higgie, 1984), but with little historical evidence to support their theses.

As Hancock has latterly argued, the loss of the third, neuter gender in Romani must have taken

place after 1000 CE whilst the ancestors of the Romany peoples were still within the orbit of New

Indo-Aryan speakers. Doman had maintained this aspect and so its speakers must have departed

230


earlier (Hancock, 2003: 2-7; Tikkanen, 2003: pc). Lomavren, Hancock goes on to argue, must have

been part of the Romani migration, but separated quite soon after leaving the environs of Indic

speakers (2003: 6). So, the current linguistic position would appear to posit at least two major

migrations leaving Indian territories, the first Domart "exodus" sometime between the early 6th

(Hancock, 2003: 2) and the close of the 9th century (Sampson, 1926: 28-29), and the Romani

departure after 1000 CE (Hancock, 2003: 4). The second of these fragmented after passing through

Persian territory (or territory predominantly influenced by the use of Persian, which may not be the

same thing), in the Armenian-speaking lands (Hancock, 2003: 6). The historical evidence for the

latter hinges around the activities of Amir (later Sultan) Mahmüd of Ghaznä's campaigns into north-

western and the central northern region of India c. 997-1030 CE. Little has been suggested about the

earlier factors involved in the forced migration of Dom, though the occasional reference has been

made to the Hephthalite Huns as a factor (Hancock, 2000: 10-11; Marushiakova & Popov, 2001: 5).

Despite this debate from the 1930's onwards, the dominant view is still one that argues that the

Gypsies left India during the 5th century. The crucial evidence for this is the legend told in the

Persian poet Firdawsi's Shahnämä. Completed around c. 1010 CE, this vast and complex epic begins

with creation and moves through the history of the pre-Islamic and Islamic dynasties of Persia. It is

built upon the structure of murder and revenge, and depicts the struggle between the nomadic

Turanians (Turks) of central Asia and the sedentary Persians, in the context of the legendary

exploits of Islamic heroes such as Rustam. The legend of the LUr and Bahrain Gur was drawn

originally by Firdawsi from Hamza al-Isfähäni's extraordinary Chronology, written in 961 CE, and

containing elements that have been conflated with the later version, especially with regard to

Bahram's Indian princess. The legend of Bahram Gur and the Uri is discussed earlier in detail, but

its "discovery" as a source for Romani history is credited to Colonel John S. Harriott in 1830 and

published in the Transactions of the Royal Asiatic Society. His suggestion that the Lüri of this story

could be identified as the ancestors of the English Romanichals has been quoted ever since, though

not always attributed correctly. I have argued in detail about the misidentification of the Lüri as

thieves, dancers and musicians, and by extension the ancestors of the Romanies. Suffice to repeat

here my argument that these were probably mercenary archers of Rajput and Jat origins, allies of

the Sassanid kings from Sind against the common enemy, the Hepthalite Huns. As such, they are

likely to have spoken an Old Indo Aryan tongue with three genders and, after their defeat and

dispersal as a decapitated military society (i. e. an early mediaeval former army with no warriors)

remained in Persia as a pariah group.

231


This problematised my original question still further; if the primary evidence of an early, single

migration has been bound to the mistaken notion of the LUr as the ancestors of the Romanies,

surely this would have encouraged more, not less scholarly attention to be focussed upon the region

and the Gypsy peoples therein? As a natural extension of such enquiries, their social and economic

position in the empires that came to dominate the region would follow. I have tried to deal with

these points through particular questions of the Gypsies in the Byzantine polity related to their

identity as purveyors of magic and sorcery and in relation to taxation status. In the context of this

point, it is striking that whilst some attention from Romani Studies scholars has been devoted to

researching the Gypsies of the Christian empire (albeit drawing upon the same basic texts), fewer

have examined the history of the Gypsies under its Ottoman successor. This is not to say that no

research has been undertaken and it has sometimes been the prerogative of Islamic studies scholars

external to the discipline that have, from various theoretical perspectives conducted research in

modem Turkey (see Altinöz 2006; 2005 for example). It is Romani Studies scholarship that has

been largely absent from this field until very recently.

Is this then a question of identity at its core, in that Romani Studies is primarily concerned with

Christian Gypsies, however nominal this aspect of identity is? The interconnection of Romani

politics and evangelical Gypsy churches might indicate that there may be some truth to such an

assertion. Prominent Pentecostal ministers in Sweden like Pastor Lars Dmitri are given high-level

political representation by that government's Integration Ministry on the basis that they speak for all

Romanies in the country, for example. Is it a question of linguistics, in that underpinning much of

Romani scholarship in general has been the work of the linguists like Hancock, Courtiade and

Mattras? The question of Eurocentrism has been raised in my discussion of Romani historiography;

as in many other disciplines over the past twenty years perhaps it is time that it was considered

more generally in Romani Studies. Is Romani Studies primarily concerned with what might be

described as European (i. e. largely non-Muslim) Gypsies? In some senses, a case may be made that

the Gypsies represent another post-colonial people, "discovered", "measured" and documented by

European scholars in the ways that tribes from Papua New Guinea or the highlands of Albania have

been, and that concepts like time and notions of civilisation underpin this work to a very great degree. It is not so surprising perhaps that the foundation of Gypsylorism and much Romani Studies

lies in a city whose academics pioneered the study of another European tribal society, the clans of

the Highlands of Scotland. Like the Friesians, Basques, Saami and Vlachs, the Gypsies may be

studied by researchers in contradistinction to the national identities surrounding them. Romani

232


history is almost always oppositional to that of the "dominant" society, integration is frequently

discussed as "assimilation" and customs and practices as "counter-cultural". This is not to suggest

that these are not fundamental aspects of the relationship between Gypsies and non-Gypsies

societies, but to question the predominant notion of these as the only basis upon which academic

knowledge can be produced. The trope of "a 1,000 years of persecution" has become the dominant

discourse of both Roma politics, and to some extent Romani Studies with an emphasis upon the

"victim hood" of the Gypsy peoples, in common with other `diaspora' peoples (Matras, 2004: 69). In

this sense, groups that do not necessarily fulfil these notions may well be perceived to be outside

the purview of Romani Studies. In this case, are we dealing with the spurious question of

"authenticity"? Is this a consequence of the Gypsylorist concern with discovering "genuine"

Gypsies? The question might be to what extent Gypsies in the Muslim lands are regarded in these

terms.

The essentialist perspective that such a concern reflects has been haunting the discipline since its

earliest days. The racial science of 19th century was predicated upon such notions and as Willems

has demonstrated, ultimately led to the policies of the Nazi state towards the Roma and Sinti

(1993). The concept of racial "hygiene" described by Burleigh & Wipperman in their analysis of the

Third Reich (Burleigh & Wipperman, 1991), relied heavily upon the categorisation of individuals

according to criteria purportedly "measuring" degrees of "authenticity". Despite the discrediting of

such notions in the aftermath of defeat, the discourse of modern anti-Gypsyism encompasses this

concern at its heart by predicating the possibility of a "true" Romani identity, bounded by

assumptions about homogeneity of populations and biological inheritance as a motor for

transmission of "culture". Its interpretation in the field of Romani Studies academic enquiry has

remained strangely unaffected by the impact of ethnicity and nationalisms. Barth's seminal

discussion in 1967 (Barth, 1969) has remained the bedrock of much anthropological and

ethnological analysis; the debt of Okely (1983) and the Amsterdam school clearly is to the crucial

concept of "cultural boundary-markers" as a basis for group identity. Their consequent analysis has

been deemed to be overly dependent upon the socially constructed elements of identity and its

concomitant concerns with marginalisation, contestation and the construction of the "Other". The

notion of the "Other" as an alternative expression of western cultural imperialism has been strongly

argued by critics of postmodernism (Sardar, 1998), and Acton has challenged its use in Romani

Studies as a "catch-all" definition (1998). The earlier expression of this as "outsider" and

"marginalised social group" have also been frequently deployed, almost interchangeably in the

233


discourse of Romani scholarship. In a sense, the scholarship of more recent Romani Studies has

itself "constructed" the Gypsies it seeks to identify and describe. The earlier archaeology of `race'

engaged in by Gypsylorists has its latter day expression in the oppositional argument surrounding

"origins" and ethnognesis. The suggestion is, at its most extreme that it is possible to pinpoint the

historical moment when the Romany peoples came into being by identifying their exact date of

"exodus" from India (Courthiade, 2004: 105-7). The trajectory of these people then follows the

familiar paths mapped out in the works of Sampson, Turner and many others and relies upon the

notion of the hermetically sealed ethnie, travelling through geographical space and historical time

but largely unaffected, or perhaps unadulterated by this process (Smith, 1986). The contact with

other similar and not so similar groups of agents to arrive as a contained and sealed unit in modem

times is intrinsic to this view. With the collapse of empires whether Byzantine, Ottoman or Soviet

the metaphorical "lid" on these units is lifted to reveal a pristine and homogeneous ethnicity and

culture. The identification of earlier units as direct relatives of these units is part of the

"rediscovery" of an ethnic history that, combined with poetics and linguistics forms the structure

out of which a "genuine" culture springs.

The crucial issue is the epistemological foundation for both these aspects of what has become

Romani history. Our knowledge of the past of the Gypsy peoples has been produced to support the

pre-existent deductions drawn from linguistics, anthropology and ethnography. It has been forced to

conform to the paradigms of "nationalist historiography' both as a consequence of primarily

locating investigation and enquiry in the ideological framework of the inevitability of the evolution

of the nation-state and by proximity to states most actively engaged in the process of "nation-

building" through re-writing history. The Balkans have been frequently identified as intrinsic to the

Gypsy people, even having been designated as "the second home of the Gypsies" (Marushiakova

and Popov, 2001: 3). This implies that there was clearly a primal "homeland" from which Romanies

are ̀ exiled' and that there is a similarly elemental relationship between the two, as there is between

Serbs, Bulgarians, Greeks and their "homelands". The notions referred to earlier about

hermetically-sealed ethnic units are absolutely crucial to the identities of differing groups in the

Balkans, interpreted through the prism of 19th century Romantic nationalism. The indivisible

relationship of "blood and soil" is another of these tropes that crucially underpins national identities

and the conflicts that spring from their competing assertions. In the question of national histories,

the extent of otherwise of the "Ottoman inheritance" is at the heart of much debate amongst Balkan

scholars. This is itself a false premise; the term derives from Turkish for "wooded mountain" and

234


the political structures, the territorial boundaries, the culture and population itself are almost

entirely the result of five hundred years of Ottoman governance (Todorova, 1997: 5-2 1). If the

Balkans are the imperial heritage of the Ottomans, then the Gypsies spread over almost the entire

peninsula are perhaps a quintessentially Ottoman people. To pursue their history along the grooves

worn by the ideological nationalism of ethnic nation-states would appear to be disastrous. Not only

is establishing a critical Romani historiography thwarted by the ineluctable need for "forging"

myth as history, without acknowledging the relationships between the two, but insistence upon an

`original homeland' and external ethnogenesis merely reinforces the identification of Gypsies as

"other" and "alien" and such a definition is deeply problematic in the context of Turkish Gypsies.

The consistent need to promote this perspective is one that is politically driven and aimed at

achieving a European minority ethnic status, and the vardo of Romani history has been harnessed to

the grai of this agenda without much consideration as to the consequences beyond the European

context. This situation is no longer tenable academically, as inevitably the need to fix firmly Gypsy

identity in a romanticised history and an ossified notion of `tradition' or `culture' will exclude,

through the imposition of selective memory those who do not fit our model of the `shared past' or

`common origins', i. e. the majority of Muslim Gypsies. This kind of collective identity, `rooted' in

the creativity of particular historic interpretations must always refer to the notions of `origin', to

points of departure for the group. It is only through the conscious construction of such that the

ability to define the group, the identity as a cultural entity, is possible. The production of history is a

deliberate act of power and creation, in this instance to support a simulacra, a false notion of

similarity and difference as applied to ourselves and others.

In the current revision of early Romany history (Hancock, 2006; 202), the establishment of Anatolia

as the crucible of the Romany peoples suggests that our "origins" lie in the middle period of the

Byzantine Empire (1040-1204 AD). The premise upon which much of this revision has taken place

is the recognition that the movement of the Oghuz Türks, and more specifically the Saldjüks, may

provide the best explanation for the entry of the Gypsy peoples into Anatolia (Hancock, 2002: 5).

The koine that became Romani more fully emerged in the lands of Asia Minor, during the period

when another language developed, that of Anatolian Turkish. The poet Yunüs Emre produced a

corpus of work that reflected the chaotic and difficult times, with his frequent references to death

and destruction (Kafadar, 1995: 5) in this newly "forged" dialect. The extraordinary situation that

grew out of the fragile and unstable Saldjük polities amounted to an increasing flood of Turkmen in

the 11`" and 12th centuries that belonged primarily to the Oghuz dialect group, and had once

235


contained the Saldjüks themselves. However, the trajectory that took them from tribal leaders to

Sultans did not restrain the other Turkmen, and these tribes pressed firmly, often aggressively

against the increasingly permeable Byzantine borders in Western Anatolia. The friction of the

eastern borders led to the debacle that was Manzikert, when the Byzantiner were defeated and the

Emperor captured, in 1071. The inrush of deeper and more penetrating raids, larger-scale migration

and frequent conflicts between Turkmen tribes resulted in the pattern of fissiparous activity that was

to remain until the Ottomans were finally to conquer the whole region in the later 15'' century CE.

The frequently mistaken identification of the Gypsies with the earlier group of Judeo-Christian

heretics known as Atsinganoi (Fraser, 1992: 45) in the Byzantine Empire has led to much confusion

in dating the appearance of the Gypsies as a distinct identity. This has additionally resulted in a

number of false assumptions about the condition of Gypsies in the Byzantine Empire that have

never been tested against the available evidence. It has been taken almost for granted that Gypsies

in Byzantium were always socially and economically marginalised, invariably nomadic and

heterodox Christian in religious belief. More recent studies (White, 2000) have not significantly

challenged these assumptions or added to our knowledge about the Gypsies in Byzantium, yet as

Hancock has pointedly remarked this period is rarely, if ever addressed by Romani Studies

scholarship (Hancock, 2002: 4). Other sources for information regarding Byzantine scholarship

remain almost entirely unused, despite the accepted significance of Soulis' paper on the subject

(1961: 141-165). Scholarship on the subject of Byzantine taxation as a source for social history is

largely absent from the works of any scholars working in the discipline (Katz, 1938: 27-32;

Andreades, 1948: 80-85; Setton, 1953: 225-259; Brand, 1968: 35-60; Neville, 2000: 47-62) outside of

very brief references (Marushiakova & Popov, 2001: 16-18).

Ottoman scholarship has, by contrast been examining the social and economic position of the

Gypsies (or Cingene and Kipti, in Ottoman Turkish) as an adjunct to analysing the complexity of

Ottoman pluralism for some time. Scholars like Altinöz (2003), Faroqhi (2001: 155-164), Simsirgil

(2001: 263-276), McCarthy (1995; 1983), Inalcik (1993) and Karpat (1985; 2001), refer implicitly

or explicitly in their work to a variety of Gypsy groups in the Empire's territories, as cingene,

Abdallar, Posha, Kiptians, Kipti, Ghurbeti, Balamoron and others. Again, the major sources for

information about the Gypsy populations of the Ottoman Empire have come from taxation and

demographic data, the very stuff of economic and social history. Romani Studies has, by and large

avoided reference to the Ottomans except in the most general terms until recently. Pott's short

article in the 1844 journal, Zeitschrift der deutscen Morganländischen Gessellschaft (3: 321-335)

236


and Paspates' "Memoir on the language of the Gypsies... " (1861), and subsequent elaboration in

the years following represent some of the earliest works in this specific field (Paspates, 1862-3). His

monumental monograph of 1870 remains a foundational work on the subject and his contribution to

the Journal of the Gypsy Lore Society (1888: 3-5) in subsequent years still provides us with much

useful, if under-utilised material regarding Ottoman Gypsies. Bond's letters of 1899 and 1901

regarding Ottoman Gypsies in Monastir (Black, 1917) are also brief, but important sources for

information about occupations (the head of the fountain-makers in 1899 and the superintendent of

the water-works in 1901 in Salonika were both Gypsies employed by Ottoman governors). Most

significantly perhaps, given the current discussions regarding the possibly non-Indic origins of the

terms Rom and Lom (Hancock, 2002: 7-12; Tikkanen, 2003: pc), Bond first posits the connection

between Rom and Rüm, or Roman (i. e. Byzantine) as the origin of the ethnonym (Black, 1917: 32).

Halliday's studies of Turkish Gypsies in Anatolia more properly falls into the area of the

ethnography of the Republic (1922: 163-189), although his research was conducted at the very close

of the Ottoman period. Modem ethnographic studies, like Kemal Silay's discussion of nomadic

identity (2002), Alford-Andrews' work (1989) or Svanberg's comparison of Greek and Turkish

Gypsies and Gypsy-like groups (1992: 40-49) continue to provide essential material about minorities

and identities, which bears upon Romani Studies, but has thus far been largely unacknowledged.

Finally, the work of Benninghaus offers important correctives to an over-reliance upon Ottoman

demographic categories and offers a further critical benchmark to assess our raw data

(Benninghaus, 2001). The most recent published discussions of Ottoman Gypsies are to be found in

Marushiakova and Popov's contribution Balkan history (2001), and the attempt to analyse the

position of the Gypsies in the Ottoman commonwealth by comparison with other imperial systems,

by Barany (2001: 50-63; 2002). The limited scope of the first (the Balkan lands of the Empire)

detracts from its usefulness, despite the fact that it is the only real attempt since Fraser's brief

section (1992: 173-178), to discuss the Gypsies of the Ottoman lands, whilst the latter are

irretrievably flawed by a reliance upon an ahistorical notion of "backwardness" vis-ä-vis European

"progression" derived from Lewis (1982), and a failure to understand the so-called millet system

(2002: 84). The impact of Ottoman scholarship about Ottoman plurality, or the "myth of the millet

system" (Braude, 1982: 69-88) and the contentious debate about the realities of dhimmi status

(Ye'or, 1996; 1998) and the false notions of pluralism (Ye'or, 1995) seems to have passed Barany

by; an all-too-frequent occurrence in Romani Studies as suggested previously.

Groups of Sindhi warriors had been present in the Persian lands since the early 5th century, often as

237


auxiliaries to the Sassanid armies of Persia or as remnants of a defeated and "decapitated' military

society, after those unsuccessful campaigns against the Hepthalite Huns by Bahrain Gur (Litvinsky,

1996: 141). Descendants of Rajput and Jhat Hindus had also been fighting with Mahmüd of Ghaznä

(997-1030 CE) after his significant campaigns into the Panjab, parts of Sindh and much of the

Ganges valley. Subsequently the Saldjüks had pursued these groups in the aftermath of their defeat

of Mahmüd's heir, Mas'ud on the steppes of Dandanaqan, 23rd May, 1040 CE. These Hindus were,

as all armies in the early mediaeval period, effectively societies on the move (Hale, 1983), with the

fighting force making-up approximately one third of the total number. The rest would have been the

armorers, grooms, smiths and metal-workers, carpenters, military engineers, servants and servitors,

tent-makers, cooks, bakers, washer-women, slaves, camp followers and children. The question of

how these groups came to be located in eastern Anatolia relates directly to this aspect of early

mediaeval warfare in the region. It is likely that from the middle of the 11th century, they were

migrating into the Armenian borderlands of eastern Asia Minor, ahead of the pressure from the

Saldjüks and their Turkmen allies. This pressure shattered the Armenian Bagratid Kingdom and

`pushed' the ancestors of the Gypsies in a variety of directions: southwards led by Armenian nobles,

north and eastwards into the Caucasus and westwards into Byzantium proper. In these liminal

borderlands, the composition of the emerging population changed with the influx of refugees and

those fleeing the chaos and turmoil of the times.

However, the pressure that Turkoman tribes people increasingly placed upon the Saldjüks

themselves, in the wake of the Byzantine defeat at Manzikert (1071 CE), led to a chaotic and

unstable situation throughout the whole region from Khorasän to central Anatolia, pushing these

groups further west. The Turkomans had followed the earlier migrations of the Oghuz or Ghuzz

Turks from Central Asia that had assaulted Ghaznäwid power with the onslaught of the

Qarakhanids, c. 999 (Luther & Bosworth, 2001: 3). Although the latter had settled in the Transoxiana

region, the Turkomans had pushed further into the Iranian plateau and eventually all of Anatolia,

undermining the integrity of the Saldjükid state, which they had helped to create. In the wake of Manzikert, and the equally disastrous defeat for the Byzantines at Myriokephalion (1176) this

process is the beginning of the ethnic and linguistic transformation of Anatolia and much of the Middle East (Luther & Bosworth, 2003: 4). In this context, the identification of the Saldjüks as the "motor" of migration for the proto-Romany populations is only partially correct; the major factor in

westward movement for many peoples at this time in this area was these other elements of the Oghuz or Ghuzz Turks, Turkomans or Turkmen. In the ethnic melange that Asia Minor became in

238


this period, the melding of the descendants of groups from Hindü origins and an admixture of ethnic

Turks, Persians, Greeks, Armenians, Laz and perhaps many other displaced and dispossessed

peoples formed the basis for the Gypsy peoples. Sharing a koine or lingua franca as a means of

communication amongst the group developed into an identifiable "tongue", in a similar process to

that of "Urdu", a Turkic word from which English derives "horde" in the remaining eastern portions

of the Ghaznävid Empire. This language developed out of the necessity to communicate military

information in a comprehensible form to groups of varying origin. The development of Arabic as a

language in the 7th to 9th centuries is a comparable phenomenon and, I have argued the experience

of this is at the heart of the process of the emergence of the Dom in the Arab lands. The language of

formal communication in the Saldjükid state was Persian, as it had been in the Ghaznävid state

previously, and it is also likely that the number of Persian loan words can be explained as a result of

this need to facilitate communication between these originally Hindi! military slaves and their

community, and the elite in the Ghaznävid period, without resorting to notions of an extended

period of time spent by the ancestors of the Gypsies in the Iranian lands. Likewise, the acquisition

of Armenian loan words during the short period of time the ancestors of the Gypsies, or proto-

Gypsies were in these lands was a very swift process, under the circumstances of conflict and war,

as I have suggested earlier. The encounter with Byzantium resulted in linguistic shifts and changes

too, as the encounter between the highly sophisticated and technologically advanced society of the

Greeks produced new words for a variety of technical areas but also altered the notion of those

within and those without the group, introducing the notion of differentiation from outsiders from the

non-Romani world.

References to Egyptians in 11 ̀h and 12'h century sources suggest that we may look for further details

about the origins in the Byzantine lands in edicts of the Emperors, extorting their people to avoid

contact with this dubious and suspect group of `non-Christians' (in terms of their membership of the

oecumene of the Orthodox Church). The differential taxation system so frequently identified with

the later Ottoman state's treatment of the Gypsies begins with the kaphalition levied upon the non-

Christian subjects of the Empire (Muslims, Jews, heterodox Christians, Mazdians, Zoroastrians and

others). The extraordinary immobility of the East Roman social and economic system, as it evolved

from the Diocletian and Constantinian reforms effectively `tied' producers to production in a legally

inherited form. The system for the agricultural sector followed this, with sons of farmers bound to

the land in the public interest. The artisans and craftspeople of the urban conurbations fared little

better; the guild system providing a means of monitoring and controlling this structure for the

239


Byzantine government, but the Egyptians of the Byzantine Empire were not incorporated into this

structure, as the Ottomans later did. Despite their differing tax-status, the ability of the central

treasury to extract revenue in the later Byzantine state was unable to efficiently tax and control

these Egyptians, resorting to special measures upon occasions for extracting revenues and

prohibiting others from consorting with them. The Ègyptians' then, established the basis of an

identity that would continue to develop throughout the region under the Byzantines and then the

Ottomans with extraordinary longevity, maintaining in some degree aspects that are present even in

contemporary Turkey and the Middle East.

In this context, the ability of the early `Gypsies' to adopt a commercially nomadic lifestyle can be

seen to be essential as a survival strategy and a result of dislocation and movement. The

possibilities available in the urban markets of Byzantium for the supply of particular goods and

services were there, and I suggest that we find commercially nomadic bands of Ègyptian' origin

travelling through the Empire and developing a clearer and more cohesive identity as Gypsies, and a

coalescing language amongst themselves, although becoming increasingly distant from Hindü

origins and more èthnically mixed' through contact with other groups in the process of migration.

The mentality of 12`h century Byzantine peasants and city-dwellers must have been deeply affected

by the insecurity of their daily lives. Subject to predation and raiding by bands of Saldjüks and

Turkmen, and the increasing collapse of any central authority left the Anatolian population

vulnerable to subjection by any rising warlord or bandit chieftain. The concern with death and

disorder, perceptible in the Armenian chronicles, and visible in the images of Mehmed Siyah Kalem

(Ekrem, 2004) and the poetry of Yünus Emre (1238-1320 CE), encapsulate the milieu in which

early Romani populations plied their trades, in common with the dervishes offering solace, sellers

of relics and artefacts associated with both Muslim and Christian saints and prophets, and Orthodox

monks and Sufis peddling spiritual security in the chaos and disorder of mediaeval Anatolia.

Wanderers and refugees from the calamities of warfare and persecution, ruination through excessive

taxation, the ravages of disease and the predation of nomads abounded on the roads and pathways

of the region during this period, many of whom may have made the choice to join stronger groups

offering protection, economic opportunity and companionship and adopt or be adopted into the

Gypsies. Here then, is the "origin of the Gypsies", that historical moment when the group emerges

distinctly from the variety that typified its make-up previously, "forged" between the hammer of the

Saldjüks and the anvil of the Byzantines, annealed and melded in the cauldron of mediaeval Anatolia and the Balkan lands.

240


This narrative that I have presented in the book is one that may plausibly present some insights into

the possible early history of the Gypsies, and their conditions in the Byzantine and Ottoman lands. I

have attempted to do so in the context of questioning notions of history and ethnicity in Romani

Studies, as these are presented elsewhere as authoritative and genuine. Here there are no such

claims only a recognition of how and why these are made, and a suggestion on my part that this is

part of the same processes. As the character of Caravaggio says to Hana in Michael Ondaatje's

novel The English Patient, "Trust me, I'm telling you a story... " (1992: 127).

241

Bibliography

Abbott, N. (1939) Rise of the North Arabian Script and its Kur'anic

Development, Chicago, Oriental Institute Publications

vol. 50

Abou-el-Haj, Rifa'at Ali (1991) Formation of the Modern State: the Ottoman Empire 16th to

18th Centuries, SUNY Series in the Social and Economic

History of the Middle East, Albany N. Y., SUNY Press

Acton, Thos. & Adrian Marsh (2008) "Glocalisation": a new phenomenon or an age-old process? Current adaptations in changes in Gypsy/Roma/Traveller

Identity in the Turkish Republic", Paper delivered to a

regular session "Globalisation and (De-)/(Re-)Construction

of Roma/Gypsy/Traveller Identities" at the 38th Congress

of the International Institute of Sociology, Budapest 26-30

June

Acton, Thos. & Adrian Marsh (2007) "The Development of Roma/Gypsy/Traveller Identity

during the candidacy for EU membership of the Turkish

Republic", Paper delivered to the Annual Conference of the Gypsy Lore Society, Manchester, 7 September

Acton, Thomas A. (forth. ) "Has Rishi gone out of style? Academic & policy

paradigms in Romani Studies", Roma, New Delhi

Acton, Thomas A. (2006a) "Romani Politics, Scholarship and the Discourse of Nation-

building. " in Adrian Marsh and Elin Strand [eds. ] Gypsies

& the Problem of Identities: Contextual, Constructed & Contested, Istanbul, Swedish Research Institute in Istanbul, Transactions no. 17, I. B. Tauris, pp. 27-37

Acton, Thomas A. (2006b) `In Memoriam', in Adrian Marsh & Elin Strand [eds. ],

Gypsies and the Problem of Identities: Contextual, Constructed and Contested, Istanbul, Swedish Research

Institute in Istanbul, Transactions no. 17, I. B. Tauris, p. 11

Acton, Thomas A. (2004) "Modernity, Culture & `Gypsies': Is there a Meta-Scientific

Method for Understanding the Representation of `Gypsies'?

And do the Dutch really Exist? " in Nicholas Saul & Susan

Tebbutt [eds. ] The Role of the Romanies: Images &

Counter Images of `Gypsies; /Romanies in European Cultures, Liverpool, Liverpool University Press, pp. 98-116

Acton, Thomas A. & Munday, G. (2001) Romani Culture & Gypsy Identity: A Companion

Volume to "Gypsy Politics & Traveller Identity", T. A.

Acton, Thomas A.

Acton, Thomas A

Acton, Thomas A.

Acton, Thomas A.

Acton, Thomas A

Acton & Gary Munday [ed. ], Hatfield, University of

Hertfordshire Press

(2000) Scholarship and the Gypsy Struggle; Commitment in

Romani Studies: A collection of papers and poems to

celebrate Donald Kenrick's 70th Year, Thomas Acton [ed. ],

Hatfield, University of Hertfordshire Press

(1998) Authenticity, Expertise, Scholarship & Politics: Conflicting

Goals in Romani Studies, Inaugural Lecture Series,

London, University of Greenwich

(1997) Gypsy Politics & Traveller Identity, T. A. Acton [ed. ]

Hatfield, University of Hertfordshire Press

(1993) "Rom migrations and the end of slavery: a rejoinder to

Fraser" Journal of the Gypsy Lore Society, 5th series, vol. 3,

no. 2, pp. 77-89

(1974) Gypsy Politics and Social Change. The Development of

Ethnic Ideology and Pressure Politics Among British

Gypsies from Victorian Reformism to Romany

Nationalism, International Library of Sociology, London &

Boston, Routledge & Keegan Paul

Acton, Thomas

& Ilona Klimova,

Akiner, Shirin

(2001)

(2004)

Aksan, Virginia H. & Daniel Goffman [eds. ]

(2007)

Aksan, Virginia (1999)

Aksan, Virginia H. (1995)

"The International Romani Union: an east European answer

to a west European question. " in Will Guy [ed] Between

Past & Future: the Roma of Central & Eastern Europe,

Hatfield, University of Hertfordshire Press, pp. 157-219

"Enduring Strangers: Mughat, Lyuli and other peripatetics

in the social fabric of central Asia" in Aparna Rao & Joseph

C. Berland [eds. ] Customary Strangers: New Perspectives

On Peripatetic Peoples in the Middle East, Africa & Asia,

Westport USA, Praeger Publishers, pp. 299-308

The Early Ottomans: Remapping the Empire, Cambridge,

Cambridge University Press

"Ottoman Military Recruitment Strategies in the Late

Eighteenth Century" in Erik Jan Zürcher [ed. ] Arming the

State; Military Conscription in the Middle East and Central

Asia, 1775-1925, London, I. B. Tauris

An Ottoman Statesman in War & Peace: Ahmed Resmi

Efendi 1700-1783, Leiden E. J. Brill

Aksu, Mustafa (2003) Türkiye'de cingene Olmak, Istanbul, Ozaan Yayincilik

al -Birüni, AN Rihän Muhammad b. Ahmad

(1867-77)

al -Birüni, Abü Rihän Muhammad b. Ahmad

Araýtirma

"Tarºkhu-1 Hind" in Sir H. M. Elliott [trans. ], John Dowson

[ed. ], The History of India; as told by its own historians, 8

vols. London, Trubner, vol. 2, chap. 1

(1879) The Chronology of Ancient Nations; an English version of

the Arabic text of the Athär-ul-bäkiya of Albirüni, C. E.

Sachau [trans. & ed. ], London, Oriental Translation Fund of

Great Britain and Ireland, W. H. Allen

Alford Andrews, Peter (2002) Ethnic Groups in the Republic of Turkey, [with Rüdiger

Benninghaus] Beihefte zum Tbinger Atlas des Vorderen

Orients Reihe (Geisteswissenschaften) Nr. 60.2, vols. I

[1994] & 2, Weisbaden, Germany

Allen, Charles (2001) The Soldier Sahibs: the Men Who Made the North West

Frontier, London, Abacus

Ali, Tariq (2003) The Clash of Fundamental isms; Crusades, Jihads and

Modernity, London, Verso

Alpman, Nazim (1994) Cingeneler, Istanbul, Ozan Yayincilik,

Altinöz, Ismail (2006) "Gypsies in the Balkans during the Ottoman period", Paper

presented to the Third International Congress on Islamic

Civilisation in the Balkans, 1-5 November, Bucharest,

Romania

(2005) "Gypsies in Ottoman Society", Paper presented to the

Gypsy Lore Society 2005 Annual Meeting and Conference

on Gypsy Studies, 9-10 September, Facultad de Filosofia,

Universidad de Granada, Spain

And, Metin (1979) "The Muharram Observances in Anatolian Turkey" in Peter

J. Chelkowski [ed. ] Taziyeh: Ritual Dance & Drama in Iran,

New York, New York University Press

Anderson, Perry (1979) Lineages of the Absolutist State, London, Verso

Andreades, Andre M. (1948) "Public Finances - Currency, Public Expenditure, Budget

and Public Revenue" in Norman Hepburn Baynes & Henry

St. Lawrence Beaufort Moss, Byzantium - An Introduction

to East Roman Civilisation, Oxford, Oxford at the

Clarendon Press

Andric, No (1994) The Bridge over the Drina, Lovett F. Edwards [trans. from

the Serbo-Croat], William H. McNeill [intro. ] London,

Harvill

Andric, No (1990) The development of spiritual life in Bosnia under the

influence of Turkish rule, I. Andric, Zelimer B. Juricic &

John F. Loud [eds. & trans. }, London, Duke University

Press

Angiolello, Giovan Maria (1873) A Short Narrative of the Life and Acts of the King Ussun

Cassino, London, Hakluyt Society

Anisimov, Evgenii (1993) The Reforms of Peter the Great; Progress Through

Coercion, [trans. ] J. T. Alexander, London, M. E. Sharpe

Anon. (1597) The Policy of the Turkish Empire, London, J. Windet for W.

S.

Anon. (1622) Newes from Turkie and Poland. Or a true and compendious

declaration of the proceedings betweene the great Turke and

his Maiestie of Poland, from the beginning of the Wanes,

until the latter end. With a relation of their daily Military

actions ... Translated out of a Latine copie, written by a

gentleman of quality, etc. printed at the Hage

Anon. (1685) A Short Relation of the most Remarkable Transactions in

Anon. [poss. Sir Charles H. Elliott]

(1914)

Archduke Habsburg, J. (1890)

Arnold, Hermann (1967)

Arnstberg, Karl Olov (1998)

Ascherson, Neal (1996)

several parts of Europe between the Christians and Turks;

including an exact diary of the siege of Buda. Written

originally by a person of honour, a volunteer in the

campagne, and done into English by P. R, London, T. B. for

Randolph Taylor

Turkey in Europe and Asia. With a map. Reprinted ...

from

the "Political Quarterly, " etc., Oxford Pamphlets, no. 38

"Gypsy Grammar", Journal of the Gypsy Lore Society, Old

Series, vol. 2, no. 3, pp. 148-160

"Some observations on Turkish & Persian Gypsies",

Journal of the Gypsy Lore Society, [3rd Series] Vo1.46,

pp. 105-122

Svenskar och zigenare - en etnologisk studie av samspelet

över en kulturell gräns, Stockholm, Carlsson

Black Sea: the Birthplace of Civilisation and Barbarism,

London, Vintage

Baer, Gabriel (1973) Fellah and Townsmen in Ottoman Egypt, Princeton,

Princeton University Press

Baihakt, Abü-l Fazi at (1862) Tarikh-i Baihaki, W. H. Morely [trans. & ed. ], Calcutta,

Bibliotecha Indica

Baihaki, Abii-l Fazl at (1867-77) "Tärikhu-s Subuktigtn" in Sir H. M. Elliott [trans. ], John

Dowson [ed. ], The History of India; as told by its own

historians, 8 vols. London, Trubner, vol. 2, chap. 3, pp. 53-

154

Bakic-Hayden, Milica (1995) "Nesting Orientalisms: the Case of Former Yugoslavia",

Slavic Review, vol. 54, no. 4 (Winter), pp. 2-18

Barakat, Ahmad (1979) Muhammad & the Jews, Delhi, Vikas Publishing House

Barany, Zoltan (2002) The East European Gypsies, Regime Change, Marginality,

and Ethnopolitics, Cambridge. Cambridge University Press

Barany, Zoltan (2001) "East European Gypsies in the Imperial Age", Ethnic &

Racial Studies, Vo1.24, No. 1, pp. 50-63

Barclay, John (1633) The Mirrour of Mindes or Barclay's Icon Animorum,

Englished by T. May, London, printed by John Beale for

Thos. Walkley

Barfield, T. (1989) The Perilous Frontier: Nomadic Empires and China, 221

BC to 1757 AD, Oxford, Blackwells

Barkey, Karen (1994) Bandits and Bureaucrats: the Ottoman Route to State

Centralization, Ithaca N. Y., Cornell University Press

Barth, F. (1969) Ethnic Groups & Boundries: the Social Organisation of

Culture Difference, [Based on a symposium held at the

University of Bergen, 23rd to 26th February 1967], Bergen

and London, Universitetsforlaget and Allen & Unwin

Barth, F. (1961) The Nomads of South Persia; the Basseri Tribe of the

Khamseh Confederacy, Oslo, Oslo University Press

Barthold, W. (V. V. ) (1945) Histoire des lures d'Asie Centrale, M. Douskis [ed. ] Paris,

Librairie d'Amerique et d'Orient Adrien-Maisonneuve

Barthold, W. (V. V. ) (1956-62) Four Studies on the History of Central Asia, V. & T.

Minorsky [trans. ] Leiden, E. J. Brill

Barthold, W. (V. V. ) (1923) Turkestan Down to the Mongol Invasion, 2nd Edition,

London, E. J. W. Gibb Memorial Series

Bartusis, Mark C. (1998) The Late Byzantine Army: Arms & Society 1204-1453, The

Middle Ages Series, Philadelphia, University of

Pennsylvania Press

Basmee Ansari, A. S. (2003) "Diät" in P. J. Bearman et al. [eds. ] The Encyclopaedia of

Islam, WebCD Edition, Leiden, Brill Academic Publishers,

vol. 2,488a

Baudier, Michel (1635) The History of the Serrail & of the Court of the Grand

Seigneur, Emperour of the Turkes, wherein is seene the

Image of the Othoman Greatnesse, a Table of Human

Passions & the Examples of Inconstant Prosperities of the

Court, E. Grimstone [trans. ] London, William Stansby for

Richard Meighen

Beasant, Sir Walter (1883) The Life and Achievements of Edward Henry Palmer [With

a portrait and with an appendix by G. F. Nicholl on

"Palmer's Work as an Oriental Scholar. "], London, John

Murray

Beck, Sam (1989) "The origins of Gypsy slavery in Romania", Dialectical

Anthropology, No. 14, pp. 53-61

Beck, Sam (1986) "Tsignani-Gypsies in Socialist Romania" Hefte für

Tsignanologie, 1-4

Belton, Brian (2005a) Questioning Gypsy Identity: Ethnic Narratives in Britain

and America, Walnut Creek CA, Alta Mira Press

Belton, Brian (2005b) "Telling Lies". part of a radio programme produced by

Evans, Simon and David Vaughan, We Still Breathe The

Air, Radio Prague, 22 January

http: //www. radio. cz/en/article/62579

Bennett, Tony (1990) Outside Literature, London, Routledge

Bercovici, Konrad (1928) The Story of the Gypsies, [illus. Charlotte Lederer] New

York, Cosmopolitan Book Corporation

Bercovici, Konrad (1932) The Incredible Balkans, New York, G. P. Putnam & Sons

Berktay, H. & Faroqhi, S. [eds. ] (1992) New Approaches to State and Peasant in Ottoman History,

Library of Peasant Studies 10, London, Frank Cass

Birkenmeier, John W. (2002) The Development of the Komnenian Army, 1081-1180,

Leiden, Brill

Bivar, A. D. H. (2008) "Hephthalites", Encyclopaedia Iranica Online Edition,

http: //tinvuri. com/4arlwr

Black, George F. (1912-13) "The Gypsies of Armenia" Journal of the Gypsy Lore

Society, [New Series], Vol. 6, no. 4, pp. 327-330

Bloch, Marc L. B. (1954) (Apologie pour l'histoire) The Historian's Craft, Peter

Putnam [trans. ], Manchester, Manchester University Press

Blunt, Sir Wilfred Scawen (1882) The Future of Islam, Paternoster Sq. London, Keegan Paul,

Trench & Co.

Bon, Ottavio (1996) The Sultän's Seraglio; an intimate portait of life at the Ottoman Court, Godfrey Goodwin (intro. & anno. from the

17th century edition of John Withers), London, Saqi Books

Bond, Rev. Lewis (1917) "The Gypsies of Monastir: two letters addressed to the late

Albert T. Sinclair, reprinted from the Bulletin of the New

York Public Library" George F. Black [ed. ] Manuscripts in

the New York Public Library, NY, New York Library

Borde [Boorde], Andrew (1547) The Fyrst Bake of the Introduction of Knowledge, the

wizyche dothe teache a man to speake parte of all maner of languages, and to know the usage and fashion of all maner

of countreys. And for to know the moste parte of all maner

of coynes of money, the whych is currant in every region. Made by Andrew Borde, of Physycke Doctor. Dedycated to

the right honorable and gracious lady Mary daughter of our soverayne Lorde Kyng Henry the eyght, London, William

Copland

Borrow, George (1981) A Journey to Eastern Europe in 1844 (thirteen letters),

Angus M. Fraser [ed. ], Edinburgh, Tragara Press

Borrow, George (1888) Romano lavo-lil; Word Book of the Romany or English

Gypsy Language, London, John Murray

Borrow, George (1874) Romano lavo-lil, word-book of the Romany, or English Gypsy language with many pieces in Gypsy, etc. London,

John Murray.

Borrow, George (1857) The Romany Rye, 2 vols. London, Nelson

Borrow, George (1851) Lavengro: the Scholar, the Gypsy, the Priest, 3 vols. London John Murray

Borrow, George (1841) The Zincali: an Account of the Gypsies of Spain, 2 vols. London, John Murray

Bosworth, Clifford Edmund (1982) "The concept of dhimma in early Islam" in Benjamin

Bruade & Bernard Lewis [eds. ] Christians & Jews in the Ottoman Empire: the functioning of a plural society, Volume I The Central Lands, New York & London, Holmes & Meier

Bosworth, Clifford Edmund (2002) "al-Zutt" in P. J. Bearman et al, [eds. ] Encyclopaedia of

Islam, New Edition, Vol. XI, Leiden, Brill, pp. 574-575

Bosworth, Clifford Edmund (2003) "Hamza al-Isfahäni" in P. J. Bearman et at [eds] The

Encylcopaedia of Islam, WebCD Edition, Leiden, Brill

Academic Publishers,

Bosworth, Clifford Edmund (2000) "al-'Utbi' in P. J. Bearman, Th. Banquis, C. E. Bosworth, et

at [eds. ] Encyclopaedia of Islam, New Edition, vol. 10,

Leiden, E. J. Brill, p. 945

Bosworth, Clifford Edmund (1999) "Saldjükids" in E. van Donzel et at [eds. ] The

Encyclopaedia of Islam CD-ROM Edition v. 1.0,

Koninklijke Brill NV, The Netherlands

Bosworth, Clifford Edmund (1990) "Karrämiyya" in P. J. Bearman, Th. Banquis, C. E.

Bosworth, et al [eds. ] Encyclopaedia of Islam, New Edition,

vol. 4, Leiden, E. J. Brill, pp. 667-669

Bosworth, Clifford Edmund (1986-) "Saldjükids" in H. A. R. Gibb, J. H. Kramer & B. Lewis et at

[eds. ] The Encyclopaedia of Islam (2nd ed. ), Leiden, E. J.

Brill

Bosworth, Clifford Edmund (1977) The Later Ghaznävids: Splendour & Decay; the Dynasty in

Afghanistan & Northern India, The University Press at

Edinburgh

Bosworth, Clifford Edmund (1973) The Ghaznävids; their Empire in Afghanistan & Eastern

Iran, 994-1040,2nd ed. Beirut, Librarie du Liban

Bowen, H. (1995) "Nizäm al-Mulk" in P. Bearman, Th. Bianquis, C. E.

Bosworth, E. van Donzel and W. P. Heinrichs [eds. ]

Encyclopaedia of Islam, New Edition, Leiden, Brill, vol. 8,

pp. 69-73

Bowen, H. (1986) " Ä1i Pas' ý h' a Tepedelenli" in P. Bearman , Th. Bianquis

C. E. Bosworth, E. van Donzel and W. P. Heinrichs [eds. ]

Encyclopaedia of Islam, New Edition, Leiden, Brill, vol. 1,

p. 398b

Brand, Charles M. (1968) Byzantium Confronts the West, Boston, Harvard University

Press

Braude, Benjamin (1982) "Foundation myths of the millet system" in Benjamin

Braude & Bernard Lewis [eds. ] Christians & Jews in the

Ottoman Empire: the functioning of a plural society,

Volume I The Central Lands, New York & London, Holmes

& Meier

Braude, Benjamin & Lewis, Bernard [eds. ]

(1982a)


Christians & Jews in the Ottoman Empire: the functioning

of a plural society, Volume I The Central Lands, New York

& London, Holmes & Meier

(1982b) Christians & Jews in the Ottoman Empire: the functioning

of a plural society, Volume 2 The Arab Lands, New York &

London, Holmes & Meier

Braude, Benjamin & Lewis, Bernard

(1982c) "Introduction" in Benjamin Bruade & Bernard Lewis [eds. ]





(1982b) "Introduction", in B. Braude & B. Lewis [eds. ] Christians

& Jews in the Ottoman Empire: the functioning of a plural

society, Volume 2 The Arab Lands, New York & London,

Holmes & Meier

Briggs, General John (1829) A History of the Rise of Mahommedan Power in India'till

the year 1612 AD, by Muhammad Käsim ibn Hindü Shah

Firishtah (Tarikh-i Firishtah), 4 vols. London (repr. Calcutta

1908 & 1966)

Briggs, General John (1830) "Essay on the Life and Writings of Ferishta", Transactions

of the Royal Asiatic Society, 92, London, pp. 341-61

Bright, Richard (1818) Travels from Vienna Through Lower Hungary, London,

Constable

Brown, John Porter (1927) The Darvishes, London, Cassell

Brown, L. Carl [ed. ] (1996) Imperial Legacy: the Ottoman Imprint on the Balkans and

the Middle East, New York, Columbia University Press

Browne, Edward G. (1902-24) A Literary History of Persia, London, Paternoster Square, T.

Fisher Unwin, 4 vols.

Brunschvig, R. (1986) "'Abd" in H. A. R. Gibb et al, [eds. ] Encyclopaedia of Islam,

New Edition, Leiden, E. J. Brill, vol. 1, pp. 24-40

Bryant, Jacob (1785) "Collections on the Zingara, or Gypsy Language, etc"

Archaeologia, or miscellaneous tracts relating to antiquity,

1-31 (1770-1845) London, Society of Antiquaries

Burleigh, Michael & Wolfgang Wipperman

(1991) The Racial State: Germany 1933 - 1945, Cambs.


Burnham, P. (1979) "Spatial mobility and political centralisation in pastoral

societies" in Equipe ecologie et anthropologie des societes

pastorales [eds. ] Pastoral production and society,

proceedings of the international meeting on nomadic

pastoralism / Production pastorale et societe : actes du

colloque international sur le pastoralisme nomade, Paris 1-3

Dec. 1976, Cambridge CUP

Burr, Malcolm (1948) "Gypsies as executioners in Turkey", Journal of the Gypsy

Lore Society, [3rd Series] vol. 27, no. 1, pp. 78-79

Burton, Sir Richard Francis (1898) The Jew, the Gypsy & el Islam, [ed. with a preface and brief

notes by W. H. Wilkins], London, Hutchinson,

Burton, Sir Richard Francis (1849) "A grammar of the Jataki-Belochki dialect", Journal of the

Bombay Branch of the Royal Asiatic Society, vol. 3, no. 1,

pp. 81-125

Busbecq, Ogier Ghislain de (1694) Four Epistles of O. G. Busbequius Concerning his Embassy

into Turkey, London, printed for J. Taylor and J. Wyat

Busbecq, Ogier Ghislain de (1968) The Turkish Letters of Ogier Ghiselin de Busbecq, Imperial

Ambassador at Constantinople 1554-1562, E. S. Foster

[trans. from the Latin 1633 ed. ], Oxford, Clarendon Press

Byhan, Dr. A. (1908-09) "Some Rumanian Gypsy Words, " Journal of the Gypsy

Lore Society, [New Series] vol. 2, pp. 45-50.

cagaptay, Soner (2003) "Citizenship policies in interwar Turkey", Nations &

Nationalism, vol. 9, no. 4, pp. 601-619

Cahen, Claude (1954-) "Baba'i" in Bernard Lewis et al. The Encyclopaedia of

Islam. A Dictionary of the Geography, Ethnography, and

Biography of the Muhammadan Peoples, New Edition,

Leiden, E. J. Brill

Cahen, Claude (1968) Pre-Ottoman Turkey; a general survey of the material and

spiritual culture and history c1071-1330, J. Jones-Williams

[trans. ], New York, Taplinger Publishing

Cahen, Claude (1983) "Djaysh" in Brenard Lewis et al [eds. ] Encyclopaedia of

Islam, New Edition, Leiden, E. J. Brill

Calvino, Italo (1978) Invisible Cities, Orlando, Florida, Harvest Books

Carmelite, Fr. Anastäs [the] (1913-14) "The Nawar or Gypsies of the East", A. Russell [trans. ],

Journal of the Gypsy Lore Society, [New Series] Vol. 7,

pp. 298-320

Celebi, Evliyä (1834-50) Narrative of Travels in Europe, Asia & Africa, in the

Seventeenth Century by Evliyä Efendi, J. von Hammer

(Purgstall)[trans. from the Turkish], vols. 1-2, London,

Oriental Translation Fund of Great Britain & Ireland

celik, Falika (2003) Gypsies (Roma) in the Orbit of Islam: the Ottoman

Experience (1450-1600), McGill University, Montreal,

unpub. MA thesis

Cezzär, Ahmed Pasha (1964) Nizämnäme- i Misir (Ottoman Egypt in the Eighteenth

Century), Stanford J. Shaw [trans. & ed. ], Harvard Middle

East Monographs No. 7, Centre for Middle East Studies,

Cambridge, Mass. Harvard University Press

Chakravarti, Ranabir (1999) "Early Medieval Bengal and the Trade in Horses: A Note",

Journal of the Economic and Social History of the Orient,

pp. 194 - 211

Chakravarti, Ranabir (1991) "Horse Trade and Piracy at Tana [Thana, Maharashtra,

India]: Gleanings from Marco Polo", Journal of the

Economic and Social History of the Orient, pp. 159 - 182

Chambers Journal (1878) Ottoman Gypsies, [poss. Francis Hindes Groome], London,

Chambers

Charanis, Peter (1953) "Economic Factors in the Decline of the Byzantine Empire"

The Journal of Economic History, vol. 13, no. 4, (Autumn),

pp. 412-24

Chelkowski, P. (2003) "Nizämi Gandjawi in P. Bearman et al [eds. ] The

Encyclopaedia of Islam, WebCD Edition, Brill Academic

Publishers

Christensen, Stephen Turk (1990) Violence & the Absolutist State: Studies in European &

Ottoman History, Köbenhavn (Copenhagen), Akademisk

Forlag

cingene (1972? ) Cingene Kartpostal, Istanbul, Baýmievi

cirakman, Asti (2002) From the "terror of the world" to the "sick man of Europe",

European Images of Ottoman Empire and Society from the

Sixteenth Century to the Nineteenth, Studies in Modern

European History, F. J. Coppa [Gen. Ed. ] New York, Peter

Lang

Clark, D. (1813) Travels in Various Countries of Europe, Asia and Africa, London

Clark, Edson L. (1878) The Races of European Turkey : Their History, condition,

and prospects, Edinburgh, William Oliphant

Clay, Christopher G. A. (2000) Gold for the Sultan; Western Bankers and Ottoman Finance,

1856 - 1881: a Contribution to Ottoman and International

Finance History, London, I. B. Tauris

Codell, J. F. & Macleod, D. S. (1998) Orientalism Transposed: the Impact of the Colonies on British Culture, Aldershot, Ashgate Press

Cohen, A. (1982) "On the realities of the millet system: Jerusalem in the

sixteenth century" in Benjamin Braude & Bernard Lewis,


of a plural society, Volume 1: The Central Lands, New York


Cole, Howard (2003) Ghaznävid Numismatics, http: //tinyurl. com/5p9fbw

Collins, Leslie J. D. (1975) "The military organisation and tactics of the Crimean

Tartars during the 16th & 17th centuries" in V. J. Parry &

M. E. Yapp [eds. ] War, Technology & Society in the Middle East, London, SSEES/Macmillan

Cook, M. A. (1971) Population Pressure in Rural Anatolia, London, Oxford

University Press

Co$gel, Metin M. (2004) Efficiency and Continuity in Public Finance: The Ottoman

System of Taxation, University of Connecticut Working

Paper, 2004-02R, http: //tinyurl. com/6823om

Cottaar, Annemarie (1998) Gypsies & Other Itinerant Groups: A Socio-Historical

Approach, [with Wim Willems & Leo Lucassen] London &

New York, St. Martin's Press

Courthiade, Marcel (2007) "Short Chronology of Rrom's History", posted to the Roma India group, 20th December,

Roma India(a yahoogr, oups. com

Crabb, James (1831) The Gipsies' Advocate, or, Observations on the Origin, Character, Manners, and Habits of The English Gipsies,

Edinburgh, Seeley

Crone, Patricia (2003) "The pay of client soldiers in the Umayyad period", Der

Islam, Band 80 Heft 2, Berlin & London, de Gruyter,

pp. 284-300

Crone, Patricia (2000) The Book of Strangers; Mediaeval Arabic Graffiti on the Theme of Nostalgia, attributed to Abu'l-Faraj al-Isfahäni, with Shmuel Moreh [trans. ], Princeton, Markus Wiener Publishers

Crone, Patricia (1986) God's Caliph: Religious Authority in the First Centuries of

Islam, [with Martin Hinds], Cambs. Cambridge University

Press

Crone, Patricia (1980) Slaves on Horses, the Evolution of the Islamic Polity,

Cambridge, Cambridge University Press

Crone, Patricia (1977) Hagarism; the Making of the Islamic World, Cambridge,

Cambidge University Press

Cunningham, Alan (1993) Anglo-Ottoman Encounters in the Age of Revolution;

Collected Essays vol. 1, Edward Ingram [ed. ], London &

Portland, Oregon, Frank Cass

Cunningham, Alan (1993) Eastern Questions in the Nineteenth Century; Collected

Essays vol. 2, Edward Ingram [ed. ], London & Portland,

Oregon, Frank Cass

Cuno, Kenneth (1999) "Muhammad Ali and the decline & revival thesis in modern

Egyptian history" in, Raouf Abbas [ed. ], Reform or

Modernization? Egypt Under Muhammad All, Cairo,

Egyptian Society of Historical Studies

Cuno, Kenneth (1992) The Pasha's Peasants; Land, Society & Economy in Lower

Egypt, 1740-1858, Cambs. CUP Middle East Library 27,

CUP

Dale, T. A. (1835) Campaigns of the Osman Sultans, Chiefly in Western Asia:

from Bayezid Ilderim to the Death of Murad the Fourth

1389-1640, from the German of Joseph von Hammer, vol. 1,

London

Dani, Ahmed Hasan (1973) Alberuni's Indica, [abridged and annotated] Islamabad,

University of Islamabad Press

Daniel, Norman (1984) Heroes and Saracens: An Interpretation of the Chansons de

Gests, Edinbrugh, Edinburgh University Press

Dankoff, R. (1982-85) Mahmüd Kashgari's Compendium of Turkic Dialects, 3

vols. Cambridge, Mass.

Dankoff, R. & Friedmann, V. A. (1991) "Earliest known text in Rumelian Romani', Journal of the

Gypsy Lore Society, [5th Series] Vol. 1, No. 1, pp. 1-20

Dankoff, R. & Kelly, J. (1989) "The languages of the world according to Evliya Qelebi",

Journal of Turkish Studies, 13, pp. 23-32

Dankoff, R.

& Robert Elsie [eds. ] (2000) Evliyä celebi in Albania and Adjacent Regions (Kosovo,

Montenegro, Ohrid) The Relevant Sections of the

Seyahatname, [trans. with a commentary and intro. ] Leiden,

E. J. Brill

Dankoff, Robert (1990) Evliya celebi in Bitlis, The Relevant Section of the

Seyahatname, [trans. with a commentary and intro. ] Leiden,

E. J. Brill

Darke, Hubert (1960) The Book of Government or Rules for kings, the Siyäsat-

näma or Siyar al-MulUk of Nizäm al-Mulk, [trans. from the

Persian] London, Routledge & Keegan Paul

Darling, Linda T. (1993) "Review: Rifa'at Abou-el-Haj's `Formation of the Modern

State: the Ottoman Empire 16th -l 8th Centuries"',

International Journal of Middle East Studies, vol. 25, no. I

(February), pp. 118-120

Darwin, Charles

& Alfred Russell Wallace (1958) Evolution by Natural Selection, [cont. Charles Darwin's

sketch of 1842 and essay of 1844, with Sir Francis

Darwin's introduction to his edition of these works, together with papers by Charles Darwin and Alfred

Wallace, read to the Linnaean Society in 1858, and

reprinted from the Society's Journal. The sketch and the essay, first published together in 1909, are entered below

under the title "The foundations of the Origin of species"], forward by Sir Gavin de Beer, Cambridge, published by the

XV International Congress of Zoology and the Linnaean

Society of London at the Cambridge University Press

Davenport, Richard Alfred (1837) The Life of Ali Pasha, of Tepeleni, Vizier of Epirus:

surnamed Asian, or the Lion, [With a portrait. ] London,

Thomas Tegg

Davies, Norman (1996) Europe: A History, Oxford, OUP

Davies, Norman (1982) God's Playground; a History of Poland in Two Volumes,

Volume 1: The Origins to 1795, Oxford, Clarendon Press

Davison, Roderic H. (1990) Essays in Ottoman & Turkish History, 1774-1923: The

Impact of the West, London, Saqi Books

Davison, Roderic H. (1990) "Turkish attitudes concerning Christian-Muslim equality in

the nineteenth century" in Roderic Davison [ed. ], Essays in Ottoman & Turkish History, 1774-1923: The Impact of the West, London, Saqi Books

Davison, Roderic H. (1982) "The millets as agents of change in the nineteenth century

Ottoman Empire" in B. Braude & B. Lewis, Christians &

Jews in the Ottoman Empire: the functioning of a plural

society, volume 2 The Arab Lands, New York & London,

Holmes & Meier

Dawes, Elizabeth A. S. [trans. ] (1918) The Alexiad of the Princess Anna Comnena, London,

Kegan, Paul, Trench, Trubner and Co., Ltd.

Dawkins, Richard McGillivray (1945) The Nature of the Chronicle of Leontios Makharias..., The

Taylorian Lecture, Oxford, The Clarendon Press

Dawkins, Richard McGillivray (1964) A Setting Forth of the Chronicle of Cyprus, beginning with

the year 1456 after Christ... The Chronicle of Giorgios

Boustronios 1456-1489, University of Melbourne Cyprus

Expedition, Melbourne University Press

de Blois, F. C. (1995) "Saldjukids; Literature in Persia and Irak" in C. E.

Bosworth et al, [eds. ] Encyclopaedia of Islam, New Edition,

vol. 8, Leiden, E. J. Brill, pp. 970-972

De Boor, C. [ed. ] (1883) Chronographia of Theophanes The Confessor, Corpus

Scriptores Historiae Byzantinae, Leipzig, Teubner (repr.

Hildesheim, 1963)

De Bruijn, J. T. (1987) "Poets & Minstrels in Early Persian Literature" Societas

Iranologica Europaea, Transition Periods in Iranian History,

Actes du Symposium de Fribourg-en-Brisgau 22-24 Mai

1985, Studia Iranica Caheir 5, Assoc. Pour I'Avancement

des Etudes traniennes

De Foletier, F. de Vaux (1971) Mille ans d'histoire des Tsiganes, Collection les grandes

etudes historiques, Paris, Fayard

de Goeje, Michael Jan (1903) "Memoire sur les migratins des Tsiganes ä travers I'Asie"

Memoires d'histoire et geographie orientales, 3, Leiden

de Goeje, Michael Jan (1886) Treatise on the Origins of the Gypsies, David Macritchie

[ed. ], with an appendix "Accounts of the Gypsies of India",

Edinburgh, Gypsy Lore Society

de Tott, Baron (1785) Sur les lures et les Tartares, Mastricht

Deb, Siddartha (2003) "Textbook Troubles", Boston Globe, 10th June,

http: //tinvurl. com/4rskrf Deringil, Selim (1998) The Well-Protected Domains: Ideology & the Legitimation

of Power in the Ottoman Empire, 1876-1909, London, I. B. Tauris

Destani, B. [ed. ] (2007) Minorities in the Middle East: Muslim Minorities in Arab

Countries, 1843 - 1973,4 vols., vol. 1,1843 - 1930,

Slough, Archive Editions

Detienne, Marcel (1986) The Creation of Mythology, Margaret Cook [trans. ],

Chicago & London, University of Chicago Press

Dickie, Matthew W. (1995) "The Fathers of the Church and the Evil Eye" in H.

Maguire [ed. ] Byzantine Magic, Dumbarton Oaks Research

Library & Collection, Washington, D. C. pp. 9-34

Digard, Jean-Pierre (2003) "Gypsies of Persia" in Ehsan Yarshater [ed. ] Encyclopaedia

Iranica, New York, Encyclopaedia Iranica Foundation,

vol. XI, pp. 412b-415b

Digard, Jean-Pierre (2003) "Gypsies of Persia" in Ehsan Yarshater [ed. ] Encyclopedia

Iranica, New York, Encyclopaedia Iranica Foundation,

Vol. XI, pp. 412b-415b

Dimitrov Str. A. (1968) Sur l'etat du Systeme des timars, des XVII - XVIII SS,

Sofia, Editions de L'Academie Bulgare des Sciences,

Insitiute d'etudes balkaniques

Dodd, Clement H. (1980) Nations in the Ottoman Empire: a Case Study in

Devolution, Hull Papers in Politics no. 18, Hull, University

of Hull Press

Dodd, E. R. (1951) The Greeks and the Irrational, Berkley & Los Angeles,

University of California Press

Domanska, Ewa (1998) "Hayden White: Beyond Irony", History & Theory, vol. 37,

no. 2 (May), pp. 173 - 181

Doornbos, Paul (1984) "Haddad", in R. V. Weekes [ed. ] Muslim Peoples: A World

Ethnographic Survey, Westport CT, Greenwood Press

Dordevic, Tihomir R. (1929) "Rumanian Gypsies", Journal of the Gypsy Lore Society,

[3rd Series] Vol. 8, pp. 7-24

Dostourian, Ara Edmond (1993) Armenia and the Crusades; the Chronicle of Matthew of Edessa, [translated from the original Armenian, with a

commentary and introduction] N. Y. & London, National

Association for Armenian Studies and Research, Armenian

Heritage Series, University Press of America

Douwes, Dick (1994) Justice & Oppression: Ottoman Rule in the Province of Damascus and the District of Hama, 1785-1841, Nijmegen,

Netherlands

(2000) The Ottomans in Syria: a History of Justice & Oppression,

[rev. ed. Justice & Oppression: Ottoman Rule in the

Province of Damascus and the District of Hama, 1785-

18411 London, I. B. Tauris

Dow, Col. Alexander (1768-72) The History of Hindostan, [trans. from the Persian], 3 vols,

Dublin, Luke White (also 3rd ed. 1792)

Duffy, John (1995) "Reactions of Two Byzantine Intellectuals to the Theory

and Practice of Magic: Michael Psellos and Michael

Italikos" in H. Maguire [ed. ] Byzantine Magic, Dumbarton

Oaks Research Library & Collection, Washington, D. C. pp.

83-97

Duijzings, Ger (1997) "The making of Egyptians in Kosovo and Macedonia" in

Cora Govers & Hans Vermeulen [eds. ] The Politics of

Ethnic Consciousness, Hants, Palgrave Macmillan

Dumont, Jean (1696) A New Voyage to the Levant; containing an Account of the

Most Remarkable Curiosities in Germany, Italy, France,

Malta & Turkey, by the Baron de Cariscroon, London, T.

Goodwin

Dumont, Paul (1982) "Jewish communities in Turkey during the last decades of

the nineteenth century in the light of the archives of the

Alliance Israelite Universelle" in B. Braude & B. Lewis,


of a plural society, volume 1, The Central Lands, New York


Dunin, Elsie (1973) "cocek as a ritual dance among Gypsy women", Makedonski Folklore, Vo1.6, pp. 193-198

Dunin, Elsie (1974) "Gypsy wedding & dance customs", Makedonski Folklore,

Vols. 7-8, pp. 317-26

Duygulu, Melih (1994) "cingene", Dünden Bugüne Istanbul Ansiklopedesi, Cilt 2,

Istanbul, Kültür Bakanligi ve Tarih Vakfi'nin Ortak

Yayinidir, pp. 514 - 516

Eich, Thomas (2003) "AbO I-Hudä, the Rifa'iya and Shi'ism in Hamidian Iraq"

Der Islam, Berlin, Walter de Gruyter vorlag, pp. 142-151

Ekrem, Isin (2004) Mehmet Siyah Kalem; Master of Humans and Demons,

Istanbul, Yapi ve Kredi Bankasi

Ellenborough, Lord (1843) "The Proclamation of the Gates, 1842", The United

Kingdom House of Commons Debate, March 9, Hansard

[HL/PO/OF, HC/OF, HAN] London, House of Lords Record Office

Elliott, Sir Charles Boileau (1838) Travels in the three great empires of Austria, Russia, and Turkey, London,

Elliott, Sir Henry Miers (1867-77) The History of India; as told by its own historians, John

Dowson [ed. ], 8 vols. Volume I: Introduction, Volume II: To

the Year A. D. 1260, Volume II: To the Year A. D. 1260,

Volume III: To the Year A. D. 1398, Volume IV: To the Year

A. D. 1450, Volume V: End of the Afghan Dynasty and the

First Thirty-Eight Years of the Reign of Akbar, Volume VI:

Akbar and Jahangir, Volume VII: From Shah-Jahan to the

Early Years of the Reign of Muhammad Shah, Volume VIII:

To End of the Muhammadan Empire in India, London,

Trubner

Elysseeff, Dr. A. (1889) "The Gypsies of Asia Minor", Journal of the Gypsy Lore

Society, [Old Series] Vol. 1, No. 5, pp. 249-50

Enoki, Kazuo (1955) "On the nationality of the Ephthalites", Memoirs of the

Research Department of the Tokyo Bunko, no. 18, Tokyo,

pp. 1-58

Eldem, Erdem, Daniel Goffman

& Bruce Masters (1999) The Ottoman City between East & West: Aleppo, Izmir &

Istanbul, Cambridge Studies in Islamic Civilization,

Cambridge, CUP

Erdem, Hakan (1996) The End of the Ottoman Slave Trade, 1861, London, Saqi

Books

Erder, L. (1975) "The measurement of pre-industrial population changes; the

Ottoman Empire from the 15th to the 17th century" Middle

East Studies, 11, p. 284

Eriksen, Helene

& Kälmän Dreisziger (2004) The Road of the Roma: from Rajasthan to Romania,

httn: //www. helene-eriksen. de/RomaProgEng. html

ERRC/hCa/EDROM (2008) We Are Here! Discriminatory Exclusion and Struggle for

Rights of Roma in Turkey, Edirne Roma Association,

European Roma Rights Centre, Helsinki Citizens'

Assembly, Istanbul

Ersanh, Büýra (2002) "The Ottoman Empire in the historiography of the Kenialist

era: a theory of fatal decline" in F. Adanir & S. Faroqhi [eds. ], The Ottomans and the Balkans: a Discussion of Historiography, Leiden, Boston, Köln, Hans Brill

Eton, William (1798) A survey of the Turkish Empire. In which are considered, I.

Its government, ... history, ... and population. 11. The state of

the provinces, ... III. The cause of the decline of Turkey, .. IV. The British commerce with Turkey, etc, London, T.

Cadell & W. Davies

European Commission (2004) The Situation of the Roma in an Enlarged European Union,

Luxembourg, Office for Official Publications of the

European Community

Evliya Efendi (1834-50) Narrative of travels in Europe, Asia & Africa in the

Seventeenth Century, vols. I&2, J. von Hammer [trans.

from the Turkish], Oriental Translation Fund of Great

Britain & Ireland, London

Fabian, Johannes (1985) Time and the Other; How Anthropology Makes its Object,

New York, Columbia University Press

Adanir, Fikret

& Suraiya Faroqhi [eds. ] (2002) The Ottomans and the Balkans: a discussion of historiography, Leiden, Boston & Köln, Hans Brill

Faroqhi, Suraiya (2002) "Coping with the central state, coping with local power: Ottoman regions and notables from the sixteenth to the

early nineteenth century" in F. Adanir & S. Faroqhi, The

Ottomans and the Balkans: a Discussion of Historiography,

Leiden, Boston & Köln, Hans Brill

Faroqhi, Suraiya (2001) "Women's work. Poverty and the privileges of guildsmen", Archiv Oreintalni, Quarterly Journal of African & Asian

Studies, vol. 69, No. 2, Oriental Institute, Academy of

Sciences of the Czech Republic, Prague, Academia

Publishing House

Faroqhi, Suraiya (1995) Subjects of the Sultan; Culture & Daily Life in the Ottoman

Empire, London, I. B. Taurus

Faroqhi, Suraiya (1995a) "Conflict, accommodation & long-term survival; the

Bektashi Order and the Ottoman State (sixteenth-

seventeenth centuries)" in A. Popov & G. Veinstein [eds. ]

Bektachiyya; etudes sur fordre des Bektachis et les groupes

relevant de Hadji Bektach, Editions Isis, Istanbul

Faroqhi, Suraiya (1992) "In search of Ottoman History" in H. Berktay & S. Faroqhi,

New Approaches to State and Peasant in Ottoman History,

Library of Peasant Studies 10, London, Frank Cass, pp. 211-

41

Faruqhi, F. M. (1998) "The Muslim Rule in India" Impact International, Vo1.28,

July http: //tinyurl. com/56m7wp

Fein, Helen (2000) "Denying the Holocaust" Lecture at London School of

Economics, Chaired by Brendan O'Leary, London

Fein, Helen (1979) Accounting for Genocide: National Responses & Jewish

Victimization during the Holocaust, New York, Free Press

Fellows [pseud. ] (1839) A Journal Written During an Excursion in Asia Minor,

Edinburgh, Murray

Fellows [pseud. ] (1841) An Account of Discoveries in Lycia, Edinburgh, Murray

Finck, Franz N. (1907) "Die Grundzge des armenisch-zigeunerischen Sprachbaus",

Journal of the Gypsy Lore Society, New Series, vol. 1, pp.

34-60

Findley, Carter V. (1982) "The acid test of Ottomanism: the acceptance of non- Muslims in the late Ottoman bureaucracy"in B. Braude &

B. Lewis [eds. ], Christians & Jews in the Ottoman Empire:

the functioning of a plural society, vol. 1 The Central

Lands, New York & London, Holmes & Meier

Fine, John V. A. (1983) The Early Medieval Balkans: A Critical Survey from the

6th to the late l 2th Centuries, Chicago, Ann Arbor

Firdawsi, Abu 11-Kasim Hasan b. Ali

(1905-25) The Shahnama of Firdawsi, Arthur George Warner and

Edmond Warner [trans], London, K. Paul Trench, Trübner

& Co. http: //erga. packhum. org/persian

Fish, Stanley (1989) "Commentary: the young and the restless" in H. Aram

Veeser [ed. ] The New Historicism, London, Routledge, pp.

303-316

Fleming, Katherine E. (1999) The Muslim Bonaparte; Diplomacy & Orientalism in All

Pasha's Greece, Princeton New Jersey, Princeton University

Press

Fletcher, J. F. (1979-80) "Turco-Mongolian monarchic tradition in the Ottoman

Empire", Harvard Ukranian Studies, 3/4,

Fögen, Marie Theres (1995) "Balsamon on Magic: from Roman Secular Law to

Byzantine Canon Law" in Henry Maguire [ed. ] Byzantine

Magic, Dumbarton Oaks Research Library & Collection,

Washington, D. C. pp. 99-115

Fonseca, Isabel (1995) Bury Me Standing: the Gypsies & their Journey, Chatto &

Windus

Foster, F. (1935-36) "Gypsy Folk Tales", Slavonic & East European Review,

Vo1.14

Foster, Sir W. [ed. ] (1931) The Travels of John Sanderson in the Levant, 1584-1602,

London, Hakluyt Society, 2nd. Series, no. LXVII

Francis, H. J. (1966) "Gypsies in Afghanistan", Journal of the Gypsy Lore

Society, [3rd Series] Vol. 45, pp. 77-84

Frankham, E. (2000) Presentation to the Romani Studies 2000 Conference, (June

- July), Greenwich University, Eltham

Fraser, Angus M. (1997) "Afterword" in Thomas A. Acton [ed. ] Gypsy politics and

Traveller identity, Hatfield, University of Hertfordshire

Press

Fraser, Angus M. (1995) The Gypsies, Peoples of Europe Series, Oxford, Blackwell

Publishing, 2nd. Ed.

Fraser, Angus M. (1992) The Gypsies, Peoples of Europe Series, Oxford, Basil

Blackwell

Fraser, Angus M. (1990) "Counterfeit Egyptians", Tsiganologische Studien, 2,43-69

Fraser, Angus M. (1951) "Marime and the Gypsies", Journal of the Gypsy Lore

Society, [3rd Series] 30,2-16 236-51

Frescobaldi, Lionardo d o Niccolö

(1818) Viaggio di L. di Niccolo Frescobaldi In Egitto E In Terra

Santa, G. Manzi [ed. ] Rome

Frye, Richard N. (1986) "Balü6istän (Balö6istän): Geography and History" in H. A.

R. Gibb, J. H. Kramers, E. Levi-Provencal, J. Schacht [eds. ]

The Encyclopaedia of Islam, New Edition, vol. 1, Leiden,

E. J. Brill, p. 1005-6

Frye, Richard N. (1974) "Historical Evidence for the Movement of People in Iran"

in György Käldy-Nagy [ed. ] The Muslim East; Studies in

Honour of Julius Germanus, Budapest, Lorand Eötvös

University

Frykman, Maja (2005) "Balkan Connections; Towards An Ethnography of Immigrants' Trans-national Practices", Towards A New

Balkan Cultural Studies, Conference 27 - 30 October,

Istanbul Bilgi University

Furnival, M. (1998) The Avars, The Dark Age Web, http: //tinyuri. com/3oa44c

Flick, J. W. (1986) "`Arabiyya - Middle Arabic" in H. A. R. Gibb et al [eds. ]

The Encyclopadia of Islam, New Edition, Leiden, E. J.

Brill, vol. 1, pp. 569-71

Gandhi, Mahatma K. (1931) Speech at the Round Table Conference, 30th November,

bttp: //tinygri. com/58adia

Gaster, Moses (1923) "Bills of sale of Gypsy slaves in Moldovia", Journal of the

Gypsy Lore Society, [3rd Series] 2,20-23

Gaster, Moses (1933) "Rumanian Gypsies in 1560", Journal of the Gypsy Lore

Society, [3rd Series] 12

Gawrych, G. W. (1987) "Sheyh Galib and Selim III; Mevlevism and the Nizam-i

Cedid", International Journal of Turkish Studies, vol. 4,

no. 1, pp. 91-114

Gent, E. B. (1812) A New Dictionary of the Terms Ancient & Modern of the

Canting Crew, in its several tribes of Gypsies, beggars,

thieves, cheats..., Philadelphia, First American [abridged

from the London edition]

Gheorghe, Nicolae. (1983) "The origins of Roma's slavery in the Rumanian

Principalities", Roma, 7/1,12-27

Gibb, Hamilton A. R. (1932) The Arab Conquest of Central Asia, London, James G.

Forlong Fund

Gibb, Hamilton A. R. (1986) "Arabiyya" in P. Bearman et al. [eds. ], Encyclopaedia of Islam, WebCD Edition, Leiden, E. J. Brill, vol. 1, pp. 561-

603

Gilbar, Gad G. (1990) Ottoman Palestine 1800-1914, Studies in Economic &

Social History, Leiden, E. J. Brill

Gillard, David (1984-5) British Documents on Foreign Affairs: Reports and Papers

from the Foreign Office Confidential Print, Part 1: from the

mid 19th Century to the 1st World War, Series B; the Near

& Middle East, vol. 1: the Ottoman Empire in the Balkans,

1856-1875, Frederick MD, University Publications of

America

Ginio, Eyal (2004) "Neither Muslims nor Zimmis: The Gypsies (Roma) in the

Ottoman State", Romani Studies 5, vol. 14, no. 2, pp. 117-

144

Ginio, Eyal (2000) "Exploring the `Other': Margaret Hasluck and the Ottoman

Gypsies" Lecture at the School of Oriental & African

Studies, University of London, 13th December

Giovo, Paolo (1546) A shorte treatise upon the Turkes chronicles ... Drawen oute

of the Italyen tong into Latyne, by F. Niger Bassianates, and

translated out of Latyne into englysh by P. Ashton. B. L,

London, Edwarde Whitchurche

Gjorgjevic, T. R. (1929) "Rumanian Gypsies in Serbia", Journal of the Gypsy Lore

Society [3rd Series] 8, pp. 7-25

Göcek, Fatma Müge (1996) Rise of the bourgeoisie, demise of empire: Ottoman

westernization and social change, New York & Oxford,

Oxford University Press

Göcek, Fatma Müge (1987) East Encounters West - France & the Ottoman Empire in

the Eighteenth Century, Studies in Middle Eastern History,

New York & London, OUP & Inst. Of Turkish Studies

Goffe, Thomas (1631a) The Raging Turk, or Bajazet the Second, London

Goffe, Thomas (1631b) The Courageous Turk, or Amurath the First, London

Gökbilgin, Tayyip (1988) "Cingeneler", Islam Ansiklopedsi, Cilt 3, Istanbul, Milli

Egitim Ba5imevi [MEB], pp. 420b - 426b

Gordon, Matthew (2006) "Interview with William Labov", Journal of English

Linguistics, no. 34, pp. 332-51

Government of the Princ ipality of Asturias

(2006) Intervention with Ethnic Minorities,

http: //tinyurt. com/6ilpkx

Graham, C. (1881) "The Avär Language" Journal of the Royal Asiatic Society,

XIII

Gramatikova, Nevena (2002) "The vildyet-name of Demir Baba as a source of the History

of Unorthodox Islam in North-Eastern Bulgaria"

Folklore/Edebiyat, 29: 1

Greenfield, Richard P. H. (1995) "A Contribution to the Study of Paleologian Magic" in H.

Maguire [ed. ] Byzantine Magic, Dumbarton Oaks Research

Library& Collection, Washington, D. C. pp. 117-53

Greliman, Heinrich M. G. (1787) Dissertation on the Gipsies, being an historical enquiry

concerning the manner of life, economy, customs and

condition of this people in Europe, and their origin, [trans.

Matthew Raper], London, P. Elmsley

Grellman, Heinrich M. G. (1783) Die Zigeuner: ein historischer Versuch über die Lebensart

und Verfassung, Sitten und Schicksale dieses Volks in

Europa, nebst ihrem Ursprung, Dessau & Leipzig,

Grierson, G. A. (1886-7) "Arabic & Persian References to Gypsies" The Indian

Antiquary, vol. 16, pp. 257-25 8

Grierson, G. A. (1889) "Doms, Jatts and the Origin of the Gypsies", Journal of the

Gypsy Lore Society, [1st Series], vol. 1, no. 1, pp. 71-76

Grosrichard, Alain (1998) The Sultan's Court; European Fantasies of the East, [trans.

Liz Heron from 1979, Structure du serail : la fiction du

despotisme asiatique dans l'Occident classique, Paris,

Editions du Seuil] , Mladen Dolar [intro. ], London, Verso

Guy, Will [David] (1975) "Ways of looking at Roms: the case of Czechoslovakia" in

Farnham Rhefisch [ed. ] Gypsies, Tinkers & Other

Travellers, London, Academic Press, pp. 201-229

Haig, T. W. (1913-36) "Sind" in M. Th. Houtsma, H. A. R. Gibb et al [eds. ] First

Encyclopaedia of Islam, Leiden, E. J. Brill, vol. IV, pp. 433-

435

Haldon, John (2000) The State & the Tributary Mode of Production, London &

New York, Verso

Haldon, John (1992) "The Ottoman State & the Question of State Autonomy:

Comparative Perspectives" in H. Berktay & S. Faroqhi,

New Approaches to State and Peasant in Ottoman History,

Library of Peasant Studies 10, London, Frank Cass, pp. 18-

108

Hale, J. (1989) "Women and War in the Visual Arts of the Renaissance" in

J. R. Mulryne & Margaret Shewring [eds. ] War, Literature

and the Arts in Sixteenth-Century Europe, London,

Macmillan Press, pp. 43-62

Hale, J. (1985) War & Society in Rennaisance Europe, 1450-1620,

London, Fontana

Hale, J. (1983) Rennaisance War Studies, London, The Hambledon Press

Hall, Edith (1989) Inventing the Barbarian; Greek self-definition through

tragedy, Oxford, Clarendon Press

Halliday, Sir William R. (1924) "The Gypsies of Turkey", Folklore Studies Ancient &

Modern, London, Methuen

Halliday, Sir William R. (1922) "Some notes upon the Gypsies of Turkey" Journal of the

Gypsy Lore Society [3rd Series], vol. 1, no. 4, pp. 163-189

Hamilton, Bernard (2001) "Dualist Heresy in the Latin Empire of Constantinople" in

Celia Hawkesworth, Muriel Heppell and Harry Norris

[eds. ] Religious Quest & National Identity in the Balkans,

Studies in Russia and East Europe, Palgrave in association

with School of Slavonic & East European Studies UCL

London, Basingstoke, Palgrave Macmillan

Hamilton, Janet & Bernard [eds. ]

(1998) Christian and Dualist Heresies in the Byzantine World

c. 650-1405: Selected Sources, [assistance with the

translation of Old Slavonic texts by Yuri Stoyanov],

Manchester Mediaeval Sources Series, Manchester

University Press, Manchester

Hancock, Ian & Vijay John (2008) On the Koine Origins of Romani: a Study of Romani Words

& Numbers of Indic Origin, unpublished research paper,

University of Texas at Austin

Hancock, Ian (2006) "On Romani Origins & Identity: Questions for Discussion"

in Adrian Marsh & Elin Strand [eds. ] Gypsies and the

Problem of Identities; Contextual, Constructed &

Contested, Transactions of the Swedish Research Institute

in Istanbul, No. 17, I. B. Tauris Istanbul & London

Hancock, lan (2006a) "On the interpretation of a word: Porrajmos as Holocaust"

in Thomas A. Acton & Michael Hayes [eds. ] Travellers,

Gypsies, Roma: The Demonisation of Difference,

Newcastle, Cambridge Scholars' Press, pp. 53-57

Hancock, Ian (2004) "Introduction", The Heroic Present Life Among the

Gypsies: the photographs and memoirs of Jan Yoors, New

York, The Monacelli Press

Hancock, Ian (2002) We Are the Romani People (Ame sam e Rromane diene),

Interface Collection, Hatfield & Paris, University of

Hertfordshire Press/Centre de recherches tsiganes

Hancock, Ian (2000) "The emergence of Roman! as a ko*fne outside of India" in

Thomas A. Acton [ed] Scholarship & the Gypsy Struggle;

Commitment in Romani Studies, Hatfield, University of

Hertfordshire Press, 1-13

Hancock, Ian (1989) Jewish Repsonses to the Porrajmos,

http: //tinyurl. com/56ubyc

Hancock, Ian (1987) The Pariah Syndrome: an account of Gypsy slavery &

persecution, Ann Arbor Michigan, Karoma Publishers

Hancock, Ian,

Siobhan Dowd & Rajko Djuric (1998)

Hanna, Nabil Sobhi (1993)

The Roads of the Roma; a PEN anthology of Gypsy writers, Hatfield, University of Hertfordshire Press

Die Ghajar, Zigeuner am Nil, sudanesische marginalien, München, Trickster Verlag

Hanna, Nabil Sobhi (1982) "Ghagar of Sett Guiranha, a study of a Gypsy community in

Egypt", Cairo Papers in Social Science 5/1, American

University in Cairo Press

Harding, D. (2001) "Forward" in 1. Zoon, On the Margins, Roma and public

services in Romania, Bulgaria and Macedonia [with a

supplement on housing in the Czech Republic], New York,

Open Society Institute

Hardy, Peter (1960) Historians of Medieval India: Studies in Indo-Muslim

Historical Writing, London, Luzac & Co.

Harriott, Col. John Staples,

Bengal Infantry (1830) "Observations of the Oriental origin of the Romanichal, or

tribe miscalled Gypsey and Bohemian (read Dec. 5 1829

and Jan. 2 1830)", Transactions of the Royal Asiatic Society

of Great Britain and Ireland, vol. 2, London, pp. 518-558

Hasluck, F. W. (1929) Christianity & Islam Under the Sultdns, Margaret Hasluck

[ed. ], vols. 1& 2, Oxford, Clarendon Press

Hasluck, Margaret M. (1948) "Firman of A. H. 1013-14 (AD 1604-5) regarding Gypsies

in the Western Balkans", Journal of the Gypsy Lore Society,

[3rd Series] 27/1,1-12

Hasluck, Margaret M. (1938) "The Gypsies of Albania", Journal of the Gypsy Lore

Society, [3rd Series] 17/2,49-61

Hathaway, Jane (1996) The Politics of Households in Ottoman Egypt, Cambs.


Heinschink, Moses (1989) "Language and Culture of the Izmir Basket-Weavers" in

Sair Balic, Milka Jauk-Pinhak, Cedo Kisic, Milan Sipka,

Rade Uhlik [eds. ] The Sarajevo Institute for Studying

National Relationships - International Scientific Assembly,

Romany Language and Culture, 9th - 11th June 1986,

Collection of authorised presentations, Sarajevo, pp. 103-

Ill

Hell, Richard (1996) The Hephthalites of Central Asia, http: //tinyurl. com/66ehvn

Helsinki Watch (1991) Destroying Ethnic Identity: the Persecution of Gypsies in

Romania, New York, Helsinki Watch Report

Henry Maguire (1995b) "Magic and the Christian Image" in H. Maguire [ed. ]

Byzantine Magic, Dumbarton Oaks Research Library &

Collection, Washington, D. C. pp. 51-71

Henry Maguire [ed. ] (1995a) "Introduction", in H. Maguire [ed. ] Byzantine Magic,

Dumbarton Oaks Research Library & Collection,

Washington, D. C.

Herriman, Anton (1892) "Little Egypt", Journal of the Gypsy Lore Society, [Old

Series] vol. 3, no. 3, pp. 152-154

Heyd, Uriel (1961) "Ulema and Westernisation in the time of Selim III and

Mahmüd II", Scripta Hierosolymitana, 9

Hibernian (1783) History of the Dispersion of the Gypsies through Europe,

Dublin, Hibernian Magazine

Higgie, B. (1984) Proto-Romanes Phonology, unpub. PhD. Thesis for the

University of Texas at Austin

Hill, Aaron (1709) A full and just account of the present state of the Ottoman

Empire in all its branches ... faithfully related from a

serious observation taken in many years travels thro' those

countries, London

Hill, Aaron (1740) A General History of the Ottoman Empire, illustrated with

useful notes and observations (after the manner of Mr

Sandys & Mr Addison), London, sold by Charles Marsh

Hindes Grooome, Francis (1891) "The Influence of the Gypsies on the Superstitions of the

English Folk", International Folklore Conference, London

Hindes Grooome, Francis (1891) "The Gipsies", in Erik Magnusson [ed. ] National Life and Thought of the various nations throughout the world. A

series of addresses by E. M. ... and others. (Sweden,

Denmark and Iceland. By E. M. ) London, T. Fisher Unwin,

p. 414

Hindes Grooome, Francis (1878) "Ottoman Gypsies" Chambers Journal, London, W. & R.

Chambers

Hitti, Philip Khüri (1916) The Origins of the Islamic State; being a translation from

the Arabic with annotations geographic and historic of the Kitäb Futüh al-Buldän of al-Imam abu-1 ̀Abbäs Ahmad ibn

Jäbir al-Balädhuri, New York, Columbia University Press

Hoffman, Michael A. (1973) "The History of Anthropology Revisited -A Byzantine

Viewpoint", American Anthropologist, New Series, vol. 75,

no. 5, pp. 1347-57

Hopkirk, Peter (1990) The Great Game: On Secret Service in High Asia, Oxford,


Horvath, Aladär (2006) "Gadjo Nation - Roma Nation", Roma Rights Quarterly

Journal of the European Roma Rights Centre, vol. 2, no. 3,

Budapest, pp. 53-57

Howorth, Henry (1889) "The Avars" Journal of the Royal Asiatic Society (New

Series) vol. XXI

Hoyland, J. (1816) A Historical Survey of the Gipsies; designed to develop The

Origin of this Singular People, and to promote The

Amelioration of their Condition, York, William Alexander

Hiibschmannovä, Milena (2004) "Origin of the Roma", Rombase, http: //tinyurl. com/5bnzli

Hughes, Rev. Thomas Smart (1830) Travels in Sicily, Greece and Albania with additions... London, 2nd ed. 2 vols.

Hurvitz, Nirmrod (2003) "From Scholarly Circles to Mass Movements: The

Formation of Legal Communities in Islamic Societies", The

American Historical Review, Vo1.108/4, October,

http: //tinyurl. com/6gfwm6

Hussey, J. M. (1950) "The Byzantine Empire in the Eleventh Century: Some

Different Interpretations", Transactions of the Royal

Historical Society, Fourth Series, vol. 32, pp. 71-85

lbbetson, W. J. (1889) "The Origins of the Gypsies", Journal of the Gypsy Lore

Society, [Old Series] vol. 2, no. 1, pp. 223-227

Inalcik, Halil (1998) "Ottoman Galata, 1453-1553", Essays in Ottoman History,

Istanbul, Istanbul, Eren Publishers

lnalcik, Haiti [ed. ] (1996) An Economic & Social History of the Ottoman Empire

1300-1914, [with Donald Quataert], Cambridge, Cambridge

University Press, 2nd ed. lnalcik, Halil (1995) "The Customs Register of Caffa, 1487 - 1490", Sources &

Studies on the Ottoman Black Sea, vol. 1, Ukranian Research Institute, Harvard University Studies in Ottoman

Documents Pertaining to Ukraine and Black Sea Countries, 2, Cambs. Mass.

Inalctk, Halil with Donald Quataert [eds. ]

(1994) An Economic & Social History of the Ottoman Empire

1300-1914, Cambs. Cambridge University Press

Inalcik, Hahl (1993) The Middle East & the Balkans under the Ottoman Empire,

Indiana University Turkish Studies 9, Bloomington, Indiana

University Press

Inalcik, Haltl (1993) "The Yürüks: their origins, expansion and economic role" in H. Inalcik [ed. ] The Middle East & the Balkans under the Ottoman Empire, Indiana University Turkish Studies 9,

Bloomington, Indiana University Press

Inalcik, Hall (1979) "Servile labour in the Ottoman Empire" in Abraham

Ascher, Tibor Halasi-Kun & Bela Kiraly [eds. ] The Mutual

Effects of the Islamic & Judeo-Christian Worlds: the East

European Pattern, Studies on Society in Change no. 3,

Brooklyn NY, Brooklyn College Press

Inalcik, Haiti (1978) The Ottoman Empire: Conquest, Organization & Economy,

London, Varorium

tnalcik, Halil (1973) The Ottoman Empire; The Classical Age, 1300-1600,

History of Civilisation, London, Weidenfeld & Nicolson

Incirlioglu, Emine Onaran (2006) "Where exactly is cin-cin Baglari? The Boundaries of a

Gypsy Neighbourhood in Ankara" in Adrian Marsh & Elin

Strand [eds. ] Gypsies & the Problem of Identities:

Contextual, Constructed & Contested, Istanbul, Svenska

forskininginstitutet i Istanbul, Transactions 17,1. B. Tauris,

pp. 191-201

Insley, Charles L. G. (2000) "Politics and Kingship in Early Eleventh Century Mercia"

Midland History, 25

Irons, W. (1979) "Political stratification among pastoral nomads" in

L'Equipe ecologie et anthropologie des societes pastorales [ed. ] Pastoral Production & Society, Cambs. CUP

Iskandar, Kai Kä'üs ibn (1951) Qäbüs Näma (A Mirror for Princes), New York, E. P. Dutton

& Co. Inc.

Ismanoglu-inan, Huri (1994) State & Peasant in the Ottoman Empire, Leiden, Hans Brill

Issawi, Charles (1982) "The transformation of the economic position of the millets in the nineteenth century", in B. Braude & B. Lewis




Jackson, Marvin R. (1985) "Comparing the Balkan demographic experience, 1860-

1970", Journal of European Economic History, XIV

Jeffries, Richard (1885) After London or Wild England, London, Cassell

Jenkins, Keith (1995) On "What is History? " From Carr and Elton to Rorty and

White, London & New York, Routledge

Jennings, Ronald C. (1993) Christians & Muslims in Ottoman Cyprus and the

Mediterranean World, 1571-1640, New York, NYUP

Jones, Sian (1997) The Archaeology of Ethnicity: Constructing Identities in the

Past & Present, London, Routledge

Kafadar, Cemal (1995) Between Two Worlds the Construction of the Ottoman

State, Berkley, University of California Press

Kalaydjieva, Luba, David Gresham & Fancesc Calafell

(2001) "Genetic Studies of the Roma (Gypsies): A Review" BMC

Medical Genetics, 2/5 http: //www. biomedccntral. com/1471-

2350/2/5

Karpat, Kemal H. (2001) The Politicization of Islam; Reconstructing Identity, State,

Faith & Community in the Late Ottoman Empire, Oxford,


Karpat, Kemal H. (1990) The Turks of Bulgaria, London, Isis

Karpat, Kemal H. (1985) Ottoman Population 1830-1914. Demographic & Social

Characteristics, Turkish & Ottoman Studies, Madison, The

University of Wisconsin Press

Karpat, Kemal H. (1983) "Population Movements in the Ottoman state in the

nineteenth century", in P. Dumont & J. Basque-Grammont

[eds. ] Economies et Societes dans l'Empire Ottoman (fin du

XVIII- debut du XX siecle) actes du colloque de

Strasbourg, l er 5 julliet 1980, Paris, Editions du Centre

National de la Recherche Scientifique, no. 601

Karpat, Kemal H. (1982) "Millets and nationality: the roots of the incongruity of

nation and state in the post-Ottoman era" in B. Braude & B.

Lewis [eds. ] Christians & Jews in the Ottoman Empire: the functioning of a plural society, vol. 1: The Central Lands,

New York & London, Holmes & Meier

Karpat, Kemal H. (1978) "Ottoman population records and the census of 1881/82-

1893" International Journal Middle East Studies, 9

Karpat, Kemal H. (1972) "The transformation of the Ottoman state, 1789-1908",

International Journal of Middle East Studies, 3: 243-281

Karpat, Kemal H. (1968) "The land regime, social structure and modernisation in the

Ottoman Empire" in W. R. Polk & R. L. Chambers [eds. ]

Beginnings of Modernization in the Middle East in the

Nineteenth Century, Chicago, University of Chicago Press

Kartlunen, Klaus (1989) "India in Early Greek Literature", Studia Orientalia 65,

Helsinki, The Finnish Oriental Society

Kaufman, T. (1984) "Explorations in proto-Gypsy phonology and

classification", Paper delivered to the 6th South Asian

Languages Analysis Round-table, University of Texas at

Austin, 25-26 May

Katz, Solomon (1938) "Some aspects of economic life in the Byzantine Empire",

The Pacific Historical Review, vol. 7, no. 1 (March), pp. 27-

39

Kazhdan, Alexander P. [ed. ] (1991) The Oxford Dictionary of Byzantium, New York, Oxford

University Press, 3 vols.

Keay, John (2000) India, a History, London, Delhi, Harper Collins

Kedrenos, Georgios (1838-9) Compendium Historiarum, [George Cedrenus] I. Bekker

[ed. ], 2 vols, Bonn, CSHB

Kenrick, Donald Simon (2004) Gypsies: From the Ganges to the Thames, [2nd. rev. ed. ],

Interface Collection, Hatfield, University of Hertfordshire

Press

Kenrick, Donald Simon (2001) "What is Domart? ", KURI, the Journal of the Dom

Research Centre, vol. 1, no. 5, http: //tinyurl. com/46keys

Kenrick, Donald Simon (2000-) "Learn Domari", KURZ the Journal of the Dom Research

Centre, vol. 1, no. 2 [and following],

http: //tinyurl. com/4czgmn

Kenrick, Donald Simon (1998) A Historical Dictionary of the Gypsies (Romanies),

Maryland, Scarecrow Press

Kenrick, Donald Simon (1993) From India to the Mediterranean; the Migration of the

Gypsies, Interface Collection, Paris, Centre de recherches

tsiganes (Universite Rene Descartes)

Kenrick, Donald Simon (1985) "The oral tradition among Romanies in Bulgaria", Lacio

Drom, Supplement 6,77-82

Kenrick, Donald Simon (1969) Morphology and Lexicon of the Romany dialect of Kotel

(Bulgaria), unpublished thesis submitted for the degree of

Ph. D. of the London University

Kenrick, Donald Simon (1966) "Notes on the Gypsies of Bulgaria", Journal of the Gypsy

Lore Society [3rd Series] 45/3-4

Keyser-Tracqui, Christine et al (2006) "Population origins in Mongolia: Genetic structure analysis

of ancient and modern DNA", Christine Keyser-Tracqui,

Eric Crubezy, Horolma Pamzsav, Tibor Varga, Bertrand

Ludes, American Journal of Physical Anthropology,

vol. 13 1. no. 2, pp. 272-281

Khalafallah, Abd-el Ghany (2003) "Arabiyya", in P. Bearman et al [eds] The Encyclopaedia of Islam, WebCD Edition, Leiden, E. J. Brill, vol.!, pp. 561-

603

Khaldun, Ibn (1987) An Arab Philosophy of History; selections from the

Prolegomena of Ibn Khaldun of Tunis (1332-1406), Charles

Issawi [ed. & trans. ], Saqi, London

Kitromilides, Pascalis M. (1994) "Imagined Communities and the origins of the National

Question in the Balkans" in P. M. Kitromilides [ed. ]

Enlightenment, Nationalism Orthodoxy, London, Varorium

Knolles, Richard (1603) The Generall Historie of the Turkes from the first beginning

of that Nation to the rising of the Othoman Familie ... Together with the lives and conquests of the Othoman

Kings and Emperours, etc, London, A. Islip

Kochanowski, Jan (1968) "Black Gypsies, White Gypsies: the Gypsies within the

perspective of Indo-European migrations" L. Slater [trans. ]

Diogenes 63, pp. 27-47

Kolukirik, Suat (2006) "Perceptions of Identity Amongst the Tarlabaýi Gypsies,

Izmir" in Adrian Marsh & Elin Strand [eds. ] Gypsies and

the Problem of Identities: Contextual, Constructed and Contested, Istanbul, Swedish Research Institute in Istanbul

Transactions no. 17, I. B. Tauris, pp. 133-140

Korobeinikov, Dmitrii (2004) "Diplomatic correspondence between Byzantium and the

Mamlük Sultänate in the fourteenth century", Al-Masaq;

Islam & the Medieval Mediterranean, vol. 16, no. 1, pp. 53-

74

Korobeinikov, Dmitrii (2003) "Orthodox Communities in Eastern Anatolia in the 13th &

14th Centuries; Part 1, the two Patriarchates - Constantinople & Antioch", Al-Masaq: Islam & the Mediaeval Mediterranean vol. 15, no. 2, pp. 197-214

Koukounas, Nikos (2004) "Synopsis Chronika: The Emperors of the 1 Ith Century"

The Internet Medieval Sourcebook,

http: //tinyurl. com/68lv6p

Krader, Lawrence (1979) "The origin of the state among nomads of Asia" in Equipe

ecologic et anthropologie des societes pastorales [eds. ]

Pastoral production and society / Production pastorale et

societe: proceedings of the international meeting on nomadic pastoralism / actes du colloque international sur le

pastoralisme nomade, Paris 1-3 Dec. 1976, Cambs.


Kramers, J. H. (1993) "Ottoman Turks - Language and Alphabets" in M. Th.

Houtsma et at [eds. ] First Encyclopaedia of Islam, 1913-

1936, Leiden, E. J. Brill

Kunt, I. Metin (1983) The Sultan's Servants: the Transformation of Ottoman

Provincial Government, 1550-1650, New York, Columbia

University Press

Kunt, I. Metin (1982) "The transformation of zimmi into askeri' in Benjamin

Braude & Bernard Lewis [eds. ] Christians & Jews in the

Ottoman Empire: the functioning of a plural society,

Volume 1 The Central Lands, New York & London, Holmes

& Meier

Kusturica, Emir (1988) Dom za veganje, [Time of the Gypsies], Forum Sarajevo

Productions

Kuznetsova, Ljalja (1998) Shaking the Dust of Ages: Gypsies and Wanderers of the

Central Asian Steppe, New York, Aperture

Köprülüzäde, Mehmed Fuad (1935) "Abdals", TUrk Halk Ansiklopidesi, Ankara, M. E. B.

Lampe, J. R. & Jackson, M. R. (1982) Balkan Economic History, 1850-1950; from Imperial

Borderlands to Developing Nations, Joint Committee on

Eastern European publication series 10, Bloomington,

Indiana University Press

Lampton, Ann K. S. (1983) "Diwan iv. Iran", Encyclopaedia of Islam, New Edition,

332-336

Lampton, Ann K. S. (1968) "The internal structure of the Saljuq Empire" in J. A. Boyle

[ed. ] The Cambridge History of Iran, The Saljuq and

Mongol Periods, vol. 5, Cambridge, CUP, pp. 203-282

Lane, George (1999) "An account of Gregory Bar Hebraeus Abu a]-Faraj nd his

relations with the Mongols of Persia", Hugoye: Journal of

Syriac Studies, Vol. 2, No. 2

Langdon, John S. (1992) Byzantium's Last Offensive in Asia Minor; the

documentary evidence for, and hagiographical lore about

John III Ducas Vatatzes' crusade against the Turks, 1222 or

1225 to 1231, (Hellenism: Ancient, Medieval, Modem 7),

New Rochelle N. Y. Aristide D. Caratzes

Lapidus, Ira M. (1988) A History of Islamic Societies, Cambridge, Cambridge

University Press

Lastivertc'i, Aristakes (1978) The History of Vardapet Arsitakes Lastivertc'i Regarding

the Sufferings Occasioned by the Foreign Peoples Living

Lawrence, Bruce B.

Lawrence, T. E.

Lawson, Fred H.

Lawson, John Cuthbert

Lazarrescu-Zoban, M. M

le Guin, Ursula

Lecomte, G.

Lee, Ken

Lecky, William E. 11

Leland, C. G.

Levi, Scott C.

Levy, A.

Levy, Juliette de Bairacli

Levy, Juliette de Bairacli

Levy, R.

Lewis, Bernard

Around Us, Robert Bedrosian [trans. & ed. ],

http: //rbedrosian. com/alint. htm

(1984) Ibn Khaldun & Islamic Ideology, Saqi, London

(1935) Seven Pillars of Wisdom, London & Toronto, Jonathan

Cape

(1999) "Persistent myths about the Muhammad Ali period" in R.

Abbas [ed. ], Reform or Modernization? Egypt under

Muhammad All, Cairo, Egyptian Society of Historical

Studies

(1910) Modern Greek Folklore and Ancient Greek Religion,

Cambs. Cambridge University Press

(1983) "Evliya celebi and the language of the rebellious efläks",

Archivum Ottomanicum, 8,307-330

(1980) Left-Hand of Darkness (Remembering Tomorrow), Harper

Collins

(2003) "al-Mu'allaqat" in P. Bearman et al [eds] The

Encyclopaedia of Islam, WebCD Edition, Leiden, Brill,

vol. 7, pp. 254a-255a

(2000) "Orientalism and Gypsylorism", Social Analysis, 44/2

(1869) History of European Morals, 2 vols. London, Longmans

(1893) The English Gipsies & Their Language [4th ed. ] Kegan

Paul, London

(2002) "Hindus Beyond theHindu Kush: Indians in the Central

Asian Slave Trade", Journal of the Royal Asiatic Society,

Series 3, vol. 12, no. 3 pp. 277-288

(1975) "Ottoman attitudes to the rise of Balkan nationalism" in

B. K. Kiraly & Fischer-Galatai [eds. ] Essays on War &

Society in East Central Europe, 1740-1920, New York,

Columbia University Press

(1953) As Gypsies Wander: being an account of life with the

Gypsies of England, Provence, Spain, Turkey & North

Africa, London, Faber & Faber

(1952) "The Gypsies of Turkey" Journal of the Gypsy Lore Society, [3rd Series] no. 31, pp. 5-13

(1951) A mirror for princes, the Qäbüs-näma of Kai Kd'üs, London, GMS

(2001) The Middle East; 2000 years of history, from the rise of

Christianity to the present day, London, Weidenfeld &

N icolson

Lewis, Bernard (1994) Race and Slavery in the Midlle East, Oxford, Oxford

University Press

Lewis, Bernard (2002) The Emergence of Modem Turkey, [3rd Edition] Studies in

Middle Eastern History, New York and Oxford, OUP

Lewis, Bernard (1977) "Ali Pasha on nationalism", Middle Eastern Studies, 10

Lewis, Bernard (1962) "Ottoman observers of Ottoman Decline", Islamic Studies,

1, pp. 71-87

Lewis, G. L. (1955) "The Secret Language of the Geygelli Yürüks" Zeki Velidi

Togan'a Armagan [A Collection Presented to Zeki Velidi],

Istanbul

Lewis, G. L. & Ch. Quelquejay (1983) "Cingäne" Encyclopaedia of Islam, 2nd Edition, vol. II, pp.

40b - 41 a, Leiden, E. J. Brill

Lewis, Raphaela (1971) Everyday Life in Ottoman Turkey, New York, Dorset Press

Lewy, Gunther (2000) The Nazi Persecution of the Gypsies, Oxford University

Press

Lindner, Rudi Paul (1981) "Nomads, horses and Huns" Past & Present 92: 14

Lindner, Rudi Paul (1982) "What was a nomadic tribe? " Comparative Studies in

Society & History 24/4: 689-711

Lindner, Rudi Paul (1983) Nomads and Ottomans in Mediaeval Anatolia, Indiana

University Uralic & Altaic Series 44, Bloomington, Indiana

Lipa, Sid (1985) "Some consequences of the influences of the European

environment on the Gypsies", Byzantine Studies, 12,

pp. 231-246

Litvinsky, B. A. (1996) "The Hepthalite Empire" in B. A. Litvinsky [ed. ] History of

Civilizations of Central Asia: the crossroads of civilizations,

AD250 to 750, Multiple History Series vol. 111, Paris,

UNESCO Publishing

Lockwood, W. G. (1985) "An introduction to Balkan Gypsies", Giessene Hefte für

Tsiganologie, 2/1,17-23

Lockwood, W. G. (1984) "Gypsies" in Richard V. Weekes [ed. ] Muslim Peoples, A

World Ethnographic Survey, 2nd ed. 2 vols. Greenwood,

Westport Press 1: 304-308

Lovell, Mary S. (1998) A rage to live: a biography of Richard and Isabel Burton,

London, Little Brown

Lucassen, Jan & E. J. Zürcher (1999) "Introduction: Conscription and resistance. The historical

context" in Erik Jan Zürcher [ed. ] Arming the State...

London, I. B. Tauris

Lucassen, Leo (2003) "The Weakness of Well Ordered Societies: Gypsies in

Western Europe, the Ottoman Empire, and India, 1400-

1914" [with Wim Willems], Review, XXVI, number 3,

2003, pp. 283-313

Lucassen, Leo (2001) "A Many-Headed Monster: the Evolution of the Passport

System in the Netherlands and Germany in the Long

Nineteenth-Century", in J. Torpey & J. Caplan [eds. ]

Documenting Individual Identity: the Development of State

Practices since the French Revolution, Princeton, Princeton

University Press, pp. 235-255

Lucassen, Leo (2000) "Gypsies in the diaspora? The pitfalls of a biblical

concept", [with Wim Willems] Histoire Sociale / Social

History, vol. 33, no. 6, pp. 251-269

Lucassen, Leo (1998) Gypsies & Other Itinerant Groups: A Socio-Historical

Approach, [with Wim Willems & Annemarie Cottaar]

London & New York, St. Martin's Press

Lucassen, Leo (1997) "Harmful Tramps: Police Professionalisation and Gypsies

in Germany, 1700-1945", Crime, Histoire & Societes, vol. 1,

no. 1, pp. 29-50

Lucassen, Leo (1993) "A Blind Spot: Migratory and Travelling Groups in Western

European Historiography", International Review of Social

History, 38,2nd August, pp. 209-235

Lusignan, Serge (1784) History of the Revolt of Ali Bey, [2nd Edition] London, J.

Phillips

Luther, Kenneth Allin [trans. ] (2001) The History of the Seljuq Turks; from the Jami' al- Tawärikh, an Ilkhanid adaptation of the Saljüq-ndma of Zahiir al-Din Nishäpüri, C. Edmund Bosworth [ed. ],

Richmond UK, Curzon Press

Lybyer, A. (1913) The Government of the Ottoman Empire in the Time of Suleiman the Magnificent, Cambridge, Mass. Harvard

University Press

Lyotard, Jean-Francois (1979) The Post-Modern Condition: A Report on Knowledge,

Manchester University Press

M. B. (1660) Leanne of a Turke, or Instructions and Advice sent from the Turkish Army at Constantinople, to the English Army at

London ... By M. B., one of the attendants of the English

Agent there, London

Macalister, Robert A. S. (1909) "A grammar and vocabulary of the language of the Nawar

or Zutt, the nomad smiths of Palestine" Journal of the

Gypsy Lore Society New Series, no. 3, pp. 120 - 126, pp.

298 - 317 continued in 1912, New Series no. 5, pp. 289 - 305, and in 1913, New Series, no. 6, pp. 161 - 240

Macalister, Robert A. S. J1914) The Language of the Nawar or Zutt, the Nomad Smiths of

Palestine, Gypsy Lore Society Monographs no. 3,

Edinburgh, Gypsy Lore Society

MacDonald, D. B. (1909) The Religious Life & Attitude in Islam, Chicago

MacKenzie, John (1995) Orientalism: History, theory and the arts, Manchester,

Manchester University Press

Macritchie, David (1886) Accounts of the Gypsies of India, collected and edited by

D. Macritchie, London, Keagan Paul and Co.

Makdisi, Ussama (2002) "Ottoman Orientalism", The American Historical Review,

vol. 107, no. 3,

http: //www. historycooperative. org/journals/ahr/ 107.3/ah030

2000768. html

Makdisi, Ussama (1997) "Reclaiming the Land of the Bible: Missionaries,

Secularism and Evangelical Modernity", The American

Historical Review, vol. 102, no. 3 (June) pp. 680-713

Makdisi, Ussama (1996) "Reconstructing the Nation-State: the Modernity of

Sectarianism in Lebanon", Middle East Report, July- September, pp. 23-26+30

Malcolm, Noel (2001) "Crypto-Christianity & Religious Amphibianism in the Ottoman Balkans: the Case of Kosovo" in Celia

Hawkesworth, Muriel Heppell & Harry Norris [eds. ],

Religious Quest & Identity in the Balkans, Basingstoke and New York, Palgrave

Malcolm, Noel (1998) Kosovo: A Short History, Basingstoke, Macmillan

Malcolm, Noel (1996) Bosnia: A Short History, 2nd Ed. Basingstoke, Macmillan

Malcolmson, Scott (1994) Empire's Edge: Travels in South-Eastern Europe, Turkey &

Central Asia, London, Faber

Malet, G. C. (Capt. ) (1855) A History of Sind, embracing the period from AD 710 to AD 1590, written in Persian at the close of the Sixteenth

Century by Mahomed Masoom, G. C. Malet and Peer

Mahomed [trans. ], R Hughes Thomas [ed. ], Bombay,

Bombay Education Society's Press,

httR: //persian. packhum. org/persi an/mai n

Mallory, J. P. & Victor H. Mair (2000) The Tarim Mummies: Ancient China & the Mystery of the

Earliest Peoples from the West, London, Thames & Hudson

Malon-McCorgray, T. K. (1996a) Iranian Huns,

http: //www. grifterrec. com/coins/huns/huns. html

Malon-McCorgray, T. K. (1996b) The Hepthalites, http: //www. silk-road. com/artl/heph. shtml

Malvinni, David (2004) The Gypsy Caravan : from Real Roma to Imaginary

Gypsies in Western Music and Film, New York, Routledge

Maqbul Ahmad, S. (1986) "Hind-The geography of India according to mediaeval

Muslim geographers" in Bernard Lewis et al [eds. ] First

Encyclopaedia of Islam, Leiden, E. J. Brill, vol. III, pp. 404-

409

Mango, Cyril & Scott, Roger [tr. ]

(1997) The Chronicle of Theophanes Confessor: Byzantine and

Near Eastern History AD 284-813 Oxford, Clarendon Press,

Marsden, William (1785) "Observation on the language of the people commonly

called Gypsies", Archaeologia Britanica, 7, London, 382-

386.

Marsh, Adrian (2008) "A brief history of the Gypsies in Turkey", We Are Here!

Discriminatory Exclusion and Struggle for Rights of Roma

in Turkey, Edirne Roma Association, European Roma

Rights Centre, Helsinki Citizens' Assembly, Istanbul, pp. 5-

20

Marsh, Adrian (2008a) "Ethnicity and Identity: who are the Gypsies? ", We Are

Here! Discriminatory Exclusion and Struggle for Rights of

Roma in Turkey, Edirne Roma Association, European Roma

Rights Centre, Helsinki Citizens' Assembly, Istanbul,

pp. 21-30

Marsh, A. & Acton, Thos. (2008) "Glocalisation": a new phenomenon or an age-old process? Current adaptations in changes in Gypsy/Roma/Traveller

Identity in the Turkish Republic", Paper delivered to a

regular session "Globalisation and (De-)/(Re-) Construction

of Roma/Gypsy/Traveller Identities" at the 38th Congress

of the International Institute of Sociology, Budapest 26-30

June

Marsh, A. & Melike Karlidag (2008)

Marsh, A. & Acton, Thos. (2007)

"A study of research literature regarding Turkish Gypsies

and the question of Gypsy identity", We Are Here!

Discriminatory Exclusion and Struggle for Rights of Roma

in Turkey, Edirne Roma Association, European Roma

Rights Centre, Helsinki Citizens' Assembly, Istanbul,

pp. 143-58

"The Development of Roma/Gypsy/Traveller Identity

during the candidacy for EU membership of the Turkish

Republic", Paper delivered to the Annual Conference of the

Gypsy Lore Society, Manchester, 7 September

Marsh, Adrian & Elin Strand [eds. ]

(2006) Gypsies & the Problem of Identities: Contextual,

Constructed & Contested, Papers presented at the First

International Romani Studies Conference in Istanbul, at the

Swedish Research Institute in Istanbul, April 10-12 2003,

Svenska forskininginstitutet i Istanbul Transactions 17,

London, I. B. Tauris

Marsh, Adrian (2005) "Gypsies and Alevis: the impossibility of Abdallar

identity? " [with Elin Strand] in Hege Irene-Markussen [ed. ]

Alevis and Alevism: Transformed Identities, Istanbul, Isis

Press, pp. 154-74

Marsh, Adrian (2003) "The Origins of the Gypsy Peoples: Identity and influence

in Romani history" KURI, Journal of the Dom Research

Centre, Fall/Winter, vol. 1, no. 9, Larnaca Cyprus, DRC,

http: //tinyurl. com/65y64b

Marsh, Adrian (2003) "Spies, deserters and undesirable persons: the Gypsies of

Cyprus, 1322-2003" [with Elin Strand] KURI, Journal of the Dom Research Centre, Spring/Summer 1/8, Larnaca

Cyprus, DRC, http: //tinyuri. com/5nwwdo

Marsh, Adrian (2000) "Gypsies and non-Gypsies in Egypt: the Zabaleen and

Ghagar Communities of Cairo" KURI -Journal of the Dom

Research Centre, Vol. 1, No. 3, Fall/Winter

http: //www. domresearcheenter. com/

Marsh, Adrian (1998) "`There are many nations yet to achieve their freedom and independence... " Ethnicity, Nationalism & the Sheikh Sa'id

Rebellion, 1925 unpublished MA thesis, University of

London

Martin, Calvin Luther (1993) In the Spirit of the Earth: Rethinking History and Time,

Baltimore, John Hopkins University Press

Martin, Calvin Luther (1987) The American Indian & the Problem of History, Calvin

Martin [ed. ] New York, Oxford University Press

Martin, Calvin Luther (1978) Keepers of the Game: Indian-Animal Relationships & the

Fur Trade, Berkley & London, University of California

Press

Marushiakova, E. & Popov, V. (2001) Gypsies in the Ottoman Empire; a contribution to the

history of the Balkans, Olga Apostolova [trans. ], Donald

Kenrick [ed. ] Hatfield & Paris, Interface Collection,

University of Hertfordshire Press & centre de recherches

tsiganes,

Marushiakova, E. & Popov, V. (1997) "Gypsies (Roma) in Bulgaria", Studien zur Tsiganologie

und Folkloristik 18, Frankfurt am Main, Peter Lang

Europafischer Verlag der Wissenschaften

Masters, Bruce (2001) Christians & Jews in the Ottoman Arab World; the Roots of

Sectarianism, Cambridge Studies in Islamic Civilisation,

Cambs. CUP

Matras, Yaron (2004) "The Role of Language in Mystifying & Demystifying

Gypsy Identity" in Nicholas Saul & Susan Tebbutt [eds. ]

The Role of the Romanies: Images & Counter Images of `Gypsies'/Romanies in European Cultures, Liverpool,

University of Liverpool Press, pp. 53 - 78

Matras, Yaron (2002a) Romani; a Linguistic Introduction, Cambs., Cambridge

University Press

Matras, Yaron (2002b) "A conflict of paradigms", Romani Studies, the continuing

Journal of the Gypsy Lore Society, 5th Series, vol. 11,

pp. 193-210

Mayall, David (2004) Gypsy Identities 1500-2000: from Egipcyans and Moon-

men to the Ethnic Romany, London, Routledge

McCarthy, Justin (2001) The Ottoman Peoples & the End of Empire, London,

Arnold

McCarthy, Justin (1997) The Ottoman Empire: an Introductory History to 1923,

Harlow, Longman

McCarthy, Justin (1995) Death & Exile, the Ethnic Cleansing of Ottoman Muslims

1821-1922, Princeton, New Jersey, The Darwin Press

McCarthy, Justin (1990) The Population of Palestine; Population Statistics of the Late Ottoman period and the Mandate, The Institute for

Palestine Studies Series, New York, Columbia University

Press

McCarthy, Justin (1987) The Ottoman Turks: an Introductory History to 1923,

Essex, Longman

McCarthy, Justin (1983) Muslims & Minorities - the Population of Ottoman

Anatolia & the End of Empire, New York & London, New

York University Press

McCarthy, Justin (1982) The Arab World, Turkey & the Balkans 1878-1914: a handbook of historical statistics, Boston, Mass. G. K. Hall,

McCarthy, Justin (1981) "The population of Ottoman Syria & Iraq" in G. Gilbar [ed. ]

Studies in Economic & Social History of the Middle East,

1800-1914, African & Asian Studies 15, University of

Haifa, UHP

McCarthy, Justin (1979) "Age, family & migration in nineteenth century Black Sea

provinces of the Ottoman Empire" International Journal of

Middle East Studies, vol. 10, no. 3, pp. 309-23

McCarthy, Justin (1976) "Ottoman Imperial & provincial salnämes" Middle East

Studies Association Bulletin, 13/2

McCarthy, Justin & C. McCarthy

(1990) Turks and Armenians: A Manual on the Armenian Question.

Washington, DC: Assembly of American Turkish

Associations

McGowan, Bruce (1981) Economic Life in Ottoman Europe: taxation, trade and the

struggle for land, 1600-1800, Studies in Modern

Capitalism, Cambridge & Paris, CUP & Editions de la

Maison des Sciences de l'Homme

McGowan, Bruce (1994) "The elites and their retinues" in HaITI Inalcik & Donald

Quataert [eds. ] An Economic & Social History of the

Ottoman Empire 1300-1914, Cambs. CUP

Melikoff, Irene (1998) "Bektashi/Kizilbas: Historical bipartition and its

consequences" in Tord Olsson, Elisabeth Özdalga &

Catharina Raudvere [eds. ], Alevi Identity; Cultural,

Religious & Social Perspectives, Istanbul, Swedish

Research Institute in Istanbul, Transactions Vol. 8

Menage, V. L. (2003) "Firdawsi (Ferdosi)" Cl. Huart, H. Masse [add. ] in P. J.

Bearman et al. [eds. ] The Encyclopaedia of Islam, WebCD

Edition, Leiden, Brill Academic Publishers

Miller, Barnette (1941) The Palace School of Mohammed the Conqueror,

Cambridge, Mass. Harvard University Press

Minorsky, V. (1986) "Lüli' in C. E. Bosworth et al. [eds] The Encyclopaedia of

Islam, New Edition, vol. V, Leiden, Brill Academic

Publishers

Mohl, M. Jules (1841) "Mudjmal al-tawarikh", Journal Asiatique (Nouveau

Journal Asiatique), Societe Asiatique, Paris, vol. 2, xii,

pp. 510-542

Mohl, M. Jules (1838-78) Le Livre des Rois, Paris, Societe Asiatique, 7 volumes

Montesino, Noma (2002) Zigenarfrägan : intervention och romantic, Lund, Lund

University Press

Morgan, David (1993) "Mongols" C. E. Bosworth et. al. [eds. ] Encyclopaedia of

Islam New Edition, vol. vii, Leiden, E. J. Brill

Morgan, David (1986) The Mongols, Oxford, Blackwells

Morley, David (2007) "Fiction", The Invisible Kings, Manchester, Carcanet Press

Moutafchieva, Vera P. (1988) Agrarian Relations in the Ottoman Empire in the 15th &

16th Centuries, Boulder & New York, East European

Monographs, Columbia University Press

Mu'ayyad, Djamal al-DTn Abü Muhammad Ilyäs ibn Yüsuf b. Zaki

[Muhammad Zilli ibn Darvish] (1951) In the Days of the Janissaries: old Turkish Life as depicted

in the "Travel Book" of Evliyä celebi, selections from

J. von Hammer (Purgstall)'s translation of the first two

volumes of The Travel Book, Alexander Palles [ed. ],

London, Hutchinson & Co.

Mujic, Muhammed A. (1952-3) "The location of Gypsies in the Yugoslav lands under

Osmali rule" Prizoli za Orientalu Filiogiu, 3/4,137-139

al-Mulk, Nizam (1954) Siyäsat-nama, M. QazTni & M. M. Chahar-dihi [eds. ],

Tehran [AH 1335]

Mundy, Martha &

Munshi, K. M. (1951) Somnath, the shrine eternal, souvenir published on the

occasion of the installation ceremony of the Linga in the

new Somanatha Temple on May 11,1951, New Delhi[? ],

Somnath Board of Trustees

Muranyi, M. (1994) "Shahaba" in B. Lewis et al. [eds. ], Encyclopaedia of Islam,

New Edition, Leiden, E. J. Brill, vol. 7, pp. 827-9

Musallam, Basim [eds. ] (2000) The Transformation of Nomadic Society in the Arab East,

University of Cambridge Oriental Publications, Cambs.


Mutafcieva, Vera P. &

Dimitrov Str. A. (1968) Sur 1'etat du Systeme des timars, des XVII - XVIII SS,

Sofia, Editions de L'Academie Bulgare des Sciences,

Insitiute d'etudes balkaniques

Narain, A. K. (1990) "Indo-Europeans in Inner Asia" in D. Sinor [ed. ]

Cambridge History of Early Inner Asia, Cambridge,

Cambridge University Press, pp. 151-76

Neville, Leonora Alice (2000) "The Marcian Treatise on Taxation and the Nature of Bureaucracy in Byzantium, " Byzantinische Forschungen,

no. 26,47-62.

Neuberger, Mary (2004) The Orient Within: Muslim Minorities and the Negotiation

of Nationhood in Modem Bulgaria, Ithaca N. Y. University

of Cornell Press

Nicholson, R. A. (1914) The Mystics of Islam, Orientalism Early Sources, vol. 8,

London, G. Bell Publishers

Nicol, Donald M. (1993) The Last Centuries of Byzantium, 1261-1453,2nd ed. Cambridge, Cambridge University Press; Ist ed. (1972)

London, Rupert Hart-Davis and N. Y. St Martin's Press

Nicolle, David (1998) Armies of the Caliphates, 862-1098, Adam Hook, Graham

Turner [illus. ], Men-at-Arms 320, Essex, Osprey Publishing

Nicolle, David (1982) Armies of Islam, 7th -1 lth Centuries, Angus McBride

[illus. ], Men-at-Arms 125, Essex, Osprey Publishing

Nizämi Gandjawi (1977) Kusraw u Shirin, Emiko Okada [trans. ], Tokyo, UNESCO

Norris, E. (1852) "Memoir on the Scythic Version of the Behistun

Inscription" Journal of the Royal Asiatic Society, XV (1st

Series) London

Norris, Harry T. (1993) Islam in the Balkans: Religion & Society between Europe

and the Arab World, London, Hurst & Co.

Obolensky, Prince Dmitri (1948) The Bogomils: a Study in Balkan Neo-Mainchaeism,

Cambs. Cambridge University Press

Okely, Judith (1983) The Traveller-Gypsies, Cambridge, Cambridge University

Press

Olsson, Tord, Özdagla, Elizabeth & Cather ine Raudvere [eds. ]

(1998) Alevi Identity: Cultural, Religious & Social Perspectives, Istanbul, Swedish Research Institute in Istanbul

Transactions Vol. 9

Ondaatje, Michael (1992) The English Patient, McClelland & Stewart, Toronto

Onsman, Andrys (2004) Defining Indigeneity in the Twenty-First Century: A Case

Study of the Frisians, New York, Edwin Mellen Press

Ortayli, Ilber (1999) "The policy of the Sublime Porte towards Nagshbandis and

other Tarigas during the Tanzimat period" in E. Özdalga

[ed. ] Naqshbandis in Western & Central Asia: Change &

Continuity, Istanbul, Swedish Research Institute in Istanbul,

Transactions Vol. 9

Ortayli, Ilber (1998) "Ottoman Modernisation & Sabataism" in Tord Olsson et.

al. [eds. ] Alevi Identity: Cultural, Religious & Social

Perspectives, Istanbul, Swedish Research Institute in

Istanbul, Transactions Vol. 9

Owen Jones, Michael (1967) "Francis Hindes Groome: "Scholar Gypsy and Gypsy

Scholar", The Journal of American Folklore, vol. 80, no.

315, January - March, pp. 71 - 80

Ozcan, Ali Refat (2006) "Marriage Among the Turkish Gypsies", The Social

Science Journal, vol. 43, pp. 461-470

Özcan, Ali Refat (1998) Turk cingeneler (Turkish Gypsies), Erzerum, Atatürk

University Press

Pagliaro, Harold E. (1972) Irrationalism in the Eighteenth Century: Papers Presented at

the Second Annual Meeting of the American Society for

Eighteenth Century Studies, Studies in Eighteenth Century

Culture Volume 2, Cleveland, American Society for

Eighteenth Century Studies

Paksoy, H. B. (1995) "Nationality or religion? Views of Central Asian Islam",

AACAR Bulletin [of the Association for the Advancement

of Central Asian Research], vol. 8/2, Autumn

Palairet, M. (1997) The Balkan Economies, c. 1800-1914; Evolution Without

Development, Cambridge Studies in Modern Economic

History, Cambs. CUP

Palmer Edward Henry, Charles Godfrey Leland and Janet Tuckey,

(1875) English-Gipsy Songs; In Rommany with metrical English

translations, Deighton, Bell & Co

Palmer, Edward H. (1867) Oriental Mysticism, Cambs. Cambridge University Press

Panaitescu, P. N. (1941) "The Gypsies in Wallachia & Moldavia; a chapter of

economic history", Journal of the Gypsy Lore Society [3rd

Series], 20,58-72

Pandey, Gyan

Paret, R.

Partner, Nancy

Partner, Nancy

Paspati, Alexander G.

Paspati, Alexander G.

Patkanoff, K. P.

Patrin

Pavlides, Andros [ed. ]

(1992) The Construction of Communalism in Colonial North India,

Delhi, Oxford University Press

(2003) "al-Kur'dn" in P. Bearman et al [eds. ] The Encyclopaedia of

Islam, WebCD Edition, Leiden, Brill, v, 400a-429b

(1998) "`Hayden White; the Form of the Content" Hayden White

Twenty-five Years On' History and Theory, vol. 3 7, no. 2,

pp. 162-172

(1997) "Hayden White (and the Content and the Form and

(1860-63)

(1870)

(1908)

(1998)

(1989)

Peake, Harold & Fleure, Herbert John

(1928)

Peeters, Paul (1922)

Pennington, A. & Levi, P. (1984)

Perkowski, Jan (1999)

Everyone Else) at the AHA", History and Theory,

Producing the Past: Making Histories Inside and Outside

the Academy vol. 36, no. 4, theme issue 36, pp. 102-1 10

"Memoir on the language of the language of the Gypsies, as

now used in the Turkish Empire", Rev. C. Hamlin D. D.

[trans. ], Journal of the American Oriental Society, vol. 7,

pp. 143-270

Etudes sur les Tchingianes ou Bohemiens sur de l'Empire

Ottoman, Constantinople

"Some words on the dialects of the Trans-Caucasian

Gypsies, Bo$a and Karaci", Journal of the Gypsy Lore

Society, New Series, vol. 1, pp. 229 - 257, continued in vol. 2, pp. 247 - 266 and 1909, vol. 2, pp. 325 - 344

Timeline of Romani History,

http: //www. geocities. com/-12atrin/timeline. htm;

htti): //www. aeocities. com/-Datrin/historv. htm

The Chronicle of Leontios Makharias (Macheras), Nicosia,

Philokypros Publications

The Steppe and the Sown, Corridors of Time Series 5,

Oxford, Clarendon Press

Traductions et traducteurs dans l'hagiographie orientale ä

l'epoque byzantine, [Extracted from Analecta Bollandiana],

Brussels, pp. 102-104

Marko the Prince: Serbo-Croat Heroic Songs, with introductory notes by Svetozar Koljevic, London

"On the legend of Demir Baba, the Iron Father" Zeitschrift

für Balkanologie, 34: 2

Petra Cech and Mozes F. Heinschink,

(1999) Sepe6ides-Romani: Grammatik, Texte und Glossar eines

türkischen Romani-Dialekts, Wiesbaden, Harrassowitz

Verlag

Petulengro [Bernard Gilliat-Smith]

(1915-16) "Report on the Gypsy Tribes of North East Bulgaria"

Journal of the Gypsy Lore Society [New Series], Vo1.9 no. 1

pp. 1-28

Pierce, Leslie H. (1993) The Imperial Harem: Women & Sovereignty in the Ottoman

Empire, Oxford, Oxford University Press

Pitcher, Donald E. (1972) A Historical Geography of the Ottoman Empire from the

earliest times to the end of the sixteenth century... Leiden,

E. J. Brill

Pitterberg, Gabriel (2003) An Ottoman Tragedy; History & Historiography at Play,

Berkeley, University of California Press

Porter-Brown, John (1868) The Dervishes, or Oriental Spiritualism, London, Trübner

Pott, August F. (1846) "Über die Sprache der Zigeuner in Syrien [About the

languages of the Gypsies in Syria]", Zeitschrift für dies

Wissenschaft der Sprache, Vol. 1, pp. 175-186

Pritsak, Omeljan (1983) The Slavs and the Avars, Centro Italiano di Studi Sull' Alto

Medioevo, Settimane di Studio, Rome

Procopius of Caesarea (1966) Anekdota: The Secret History, G A. Williamson [trans.

With an introduction], Harmondsworth, Penguin Classics

Prosper, Mdrimee (1978) cingene, Istanbul, Altin Kitaplar Yaymevi

Quataert, Donald (2000) The Ottoman Empire, 1700-1922, New Approaches to

European History 17, Cambs. CUP

Quataert, Donald (1994) An Economic and Social History of the Ottoman Empire,

1300-1914, [eds. with Halil Inalcik], Cambridge,

Cambridge University Press, [repr. 1996]

Quataert, Donald (1983) Social Disintegration and Popular Resistance in the

Ottoman Empire, 1881-1908: Reactions to European

Economic Penetration, New York & London, New York

University Press

Rabin, 1. (2003) "Arabiyya", in P. Bearman et al [eds] The Encyclopaedia of

Islam, WebCD Edition, Leiden, E. J. Brill, vol. 1, pp. 561-

603

Rath, Jan (1991) Minorisation: the Social Construction of Ethnic Minorities,

Amsterdam, SUA

Ravenstein, E. G. (1876) "Distribution of the population in the part of Europe

overrun by Turks", The Geographical Magazine, 3

Redhouse, Sir J. W. (1884) A Simplified Grammar of the Ottoman-Turkish Language,

Trübner's Collection of Simplified Grammars

Redhouse, Sir J. W. (1909) Turkish Vade-Mecun of Ottoman Colloquial Language [4th

ed. ], Kegan, Paul & Co.

Reid, James J. (2000) Crisis of the Ottoman Empire: prelude to collapse 1839-

1878, Stuttgart, Franz Steiner Verlag

Reynolds, Dwight Fletcher (1995) Heroic Poets, Poetic Heroes: the Ethnography of

Performance in an Arabic Oral Epic Tradition, New York,

Cornell University Press

Rice, Tamara Talbot (1961) The Seljuks in Asia Minor, Ancient Peoples & Places

Series, London, Thames & Hudson

Rishi, W. R. (1996) Roma - The Panjabi Emigrants in Europe, Central and

Middle Asia, the USSR, and the Americas, Patiala, India,

Punjabi University

Rishi, W. R. (1995) "An Appeal from Dr. W. R. Rishi: Roma - Descendants of Warrior Classes of India", Roma, no. 42, January, Delhi

Rishi, W. R. (1989) "History of [the] Romano movement, their language and

culture", in Sair Balic, Milka Jauk-Pinhak, Cedo Kisi6,

Milan Sipka, Rade Uhlik [eds. ] The Sarajevo Institute for

Studying National Relationships - International Scientific

Assembly, Romany Language and Culture, 9th - 11th June

1986, Collection of authorised presentations, Sarajevo,

pp. l-10

Rishi, W. R. (1983) "On the Indian love of buffalo milk", Roma, Delhi

Rizvi, S. A. A. & J. Burton-Page (1986) "Narb [warfare] - India" in Bernard Lewis et al, [eds. ]

Encyclopaedia of Islam, New Edition, Leiden, E. J. Brill,

vol. 111, pp. ] 98-203

Robinson, Chase F. (2003) Islamic Historiography, Themes in Islamic History,


Roma Virtual Network (2007) "MAHMCD'S DAY: A DAY OF HIGH SIGNIFICANCE

IN RROMANI HISTORY" and responses, 20th December,

Roma India@yahoogroups. com

Roseman, Curtis C. (1971) "Migration as a Temporal and Spatial Process", Annals of the Association of American Geographers, vol. 61, no. 3

(September), pp. 589-598

Roux, Jean-Paul (1987) "The Tahtaci of Anatolia" in Aparna Rao [ed. ] The Other

Nomads, (preface Fredrik Barth), Köln & Wien, Böhlau

Verlag

Rowbotham, Sheila (1972) Hidden From History: 300 Years of Women's Oppression &

the Fight Against It, London, Pluto Press

Runciman, Sir Stephen (1947) The Mediaeval Manichee: A Study of the Christian Dualist

Heresy, Cambs. Cambridge University Press

Sabatino Lopez, Robert (1951) "The Dollar of the Middle Ages", The Journal of Economic

History, pp. 209-34

Sabatino Lopez, Robert (1945) "Silk Industry in the Byzantine Empire", Speculum: A

Journal of Medieval Studies, Medieval Academy of America, pp. Vol. 20, no. 1 (Jan. ) pp. 1-42

Sachau, Eduard [ed. ] (1910) Alberuni's India; an account of the religion, philosophy, literature, geography, chronology, astronomy, customs, laws

and astrology of India about AD 1030, Trübner's Oriental

Series, London, Kegan Paul, Trench, Trübner & Co.

Said, Edward (2004) "US: a disputed history of identity", Le Monde

Diplomatique, September,

http: //mondediplo. com/2004/09/04histoa

Said, Edward (1978) Orientalism, London & Henley, Routledge & Kegan Paul

Sampson, John (1927) " Notes on Professor R. L. Turner's the Position of Romani

in Indo-Aryan" Journal of the Gypsy Lore Society [3rd

Series] vol. 6,57-68.

Sampson, John (1923) "On the origin and early migrations of the Gypsies",

Journal of the Gypsy Lore Society [3d Series] vol. 2,

pp. 156-69

Sampson, John (1907) "Gypsy Language and Origin" Journal of the Gypsy Lore

Society, [New Series], vol. 1, p. 422

Sardar, Ziauddin (1998) Orientalism, Buckingham, Open University Press

Sardar, Ziauddin (1998) Postmodernism & the Other, London, Pluto Press

Sardar, Ziauddin (1992) "Terminator 2: Modernity, Postmodernity and the `Other"',

Futures, no. 24, vol. 5, pp. 493-506 ,

Sardar, Ziauddin (1979) "Science in Turkey: Choosing the Wrong Priorities",

Nature, no. 282, pp. 668-70 Sardar, Ziauddin (1977) Science, Technology & Development in the Muslim World,

London, Croom Helm

$anlier, Sinan (2006) "Romani-Türkce Sözlük (Paspatis Derlemsei Hazirlayan)"

[Romani-Turkish Dictionary, from Paspates, collected and

prepared by Sinan Sanlier], Insanokur. Org,

http: //www. insanokur. orp, /? p=263;

"Romani-Türkce Sözlük (Grellman Derlemsei Hazirlayan)"

[Romani-Turkish Dictionary, from Grellman, collected and

prepared by Sinan $anlier], Insanokur. Org,

http: //www. insanokur. orgj? p=262

Sariönder, Refika (2000) Publicising religion, differentiation & de-differentiation:

Alevis between Berlin and Anatolia, Berlin, Franz Steiner

Verlag

Savarkar, V. D. "Veer" (1949) Hindutva: Who Is A Hindü? Poona, S. P. Gokhale; also

available at http: //www. Hindüsarise. com/

Savvides, Alexes G. K. (1981) Byzantium in the Near East: its Relations with the Seljuq

Sultinate of Rum in Asia Minor, the Armenians of Cilicia,

and the Mongols, A. D. 1192-1237, Thessalonike, Kentron

Vyzantino L_ n Ereuno r' n

Schaff, Philip (2005) The Seven Ecumenical Councils, Grand Rapids, MI, USA,

Christian Classics Ethereal Library

Schäuble, Michaela (2006) "'Imagined suicide': self-sacrifice and the making of heroes

in post-war Croatia" Anthropology Matters Journal, vol. 8,

no. 1, pp. 1-14

http: //www. anthropologymatters. com/j ournaU200 6-

1 /index. htm

Schivelbusch, Wolfgang (2004) The Culture of Defeat On National Trauma, Mourning, and Recovery, Cambridge, Granta

Seeman, Sonia Tamar (2006) "Presenting "Gypsy", Re-Presenting Roman: Towards an Archaeology of Aesthetic Production and Social Identity",

Music & Anthropology; the Journal of Musical

Anthropology of the Mediterranean, Number 11,

http: //levi. provincia. venezia. it/ma/index/number 11 /seeman/

see_0. htm

Seeman, Sonia Tamar (2002) "You're Roman! " Music and Identity in Turkish Roman

Communities, unpublished dissertation, University of

California Los Angeles

Sesic, Milena Dragicevic (2005) "Balkan crossroads: for a new ethics in cultural policy

making and international relations", A Paper presented at

Towards A New Balkan Cultural Studies Conference,

Istanbul Bilgi University, 27-30 October

Shakir, M. H. (trans. ) (1997) The Holy Qur'an, University of Virginia Library, Electronic

Text Centre, http: //etext. lib. virginia. edu/koran. html

Shankland, David (1998) "Anthropology and Ethnicity: the place of ethnography in

the New Alevi Movement" in Tord Olsson et. al. [eds. ] Alevi

Identity: Cultural, Religious & Social Perspectives,

Istanbul, Swedish Research Institute in Istanbul,

Transactions Vol. 9

Shaw, Ezel Kural & Colin J. Heywood

(1972) English & Continental Views of the Ottoman Empire,

1500-1800, Papers Read at a Clark Library Seminar

January 24,1970 (with an intro. By G. E. von Grunebaum)

William Andrews Clark Memorial Library, University of

California, Los Angeles

Shaw, Stanford Jay (1978) "The Ottoman census system and population, 1831-1914"

International Journal Middle East Studies, 9

Shaw, Stanford Jay (1977) History of the Ottoman Empire & Modern Turkey, Volume

1: Empire of the Gazis: the rise & decline of the Ottoman

Empire, 1280-1808, [with Ezel Kural Shaw] Cambs. CUP

Shaw, Stanford Jay (1977) History of the Ottoman Empire & Modern Turkey, Volume

2: Reform, Revolution & Republic: the rise of modern

Turkey, 1808-1975, [with Ezel Kural Shaw] Cambs. CUP

Shaw, Stanford Jay (1971) Between Old & New, the Ottoman Empire under Sultan

Selim 111,1789-1807, Cambridge, Mass. Harvard

University Press

Simýirgil, Ahmet (2001) "Economic Life in Nove Zamky (Uyvar) during the

Ottoman period, 1663-1683" in Turkish Review of Balkan

Studies, Annual 2001,6, Foundation for Middle East &

Balkan Studies (OBIV), Istanbul, Isis

Sinclair, Albert Thomas (1917) American Gypsies, edited from manuscripts in the

collection of New York Public Library by George F. Black

(reprinted from the Bulletin of the New York Public

Library)

Sinor, Denis (1990) "The establishment and dissolution of the Türk empire" in

Denis Sinor [ed], Cambridge History of Early Inner Asia,

Cambs. Cambridge University Press, pp. 285-316

Smith, Athony D. (1986) The Ethnic Origins of Nations, Oxford, Basil Blackwell

Soucek, Suat (2000) A History of Inner Asia, Cambridge CUP

Soulis, George C. (1959) "A note on the taxation of the Balkan Gypsies in the

seventeenth century", Journal of the Gypsy Lore Society

[3rd Series], Vol. 38, No. 3-4, pp. 154-156

Soulis, George C. (1961) "The Gypsies in the Byzantine Empire and the Balkans in

the late Middle Ages", Dumbarton Oaks Papers, 15,

Dumbarton Oaks Research Library & Collection, pp. 141-

165

Speck, Paul (2003) Understanding Byzantium; studies in Byzantine historical

sources, Sarolta Takäcs [ed] Aldershot, Variorum Collected

Studies Series

Spiegel, Gabrielle M. [ed. ] (2005) Practicing History: New Directions in Historical Writing

After the Linguistic Turn, London, Routledge

Spiegel, Gabrielle M. (1999) The Past As Text: the Theory and Practice of Mediaeval

Historiography, Baltimore & London, John Hopkins

University Press

Spiegel, Gabrielle M. (1993) Romanicing the Past: the Rise of Vernacular Prose

Historiography in Thirteenth Century France, Berkeley,

University of California Press

Spiegel, Gabrielle M. (1990) "History, Historicism and the Social Logic of the Text of the Middle Ages", Speculum, 65, pp. 59-86

Stewart, Angus Donal (2001) The Armenian Kingdom and the Mamluks: War and

Diplomacy During the Reigns of Het'Um 11 (1289-1307),

Leiden, Brill

Stewart, Michael (2002a) "Review", Romani Studies: the Continuing Journal of the

Gypsy Lore Society, vol. 11, no. 2, pp. 185 - 190

Stewart, Michael (2002b) "Hungary's post-communist Roma policy: a model or a

muddle? " British Association for Central and Eastern

Europe LECTURE, Wednesday 13th March 2002, London

Stewart, Michael (1997) The Time of the Gypsies, Boulder, Colorado, Westview

Press

Stewart, Michael (1997) "The puzzle of Roma persistence: group identity without a

nation" in T. A. Acton & G. Mundy [eds. ] Romani Culture

and Gypsy Identity, Hatfield, Hertfordshire University Press Stillman, N. A. (1986) "Khil'a" in C. E. Bosworth et al [eds. ] The Encyclopaedia of

Islam New Edition, Vol. V, Leiden, E. J. Brill, pp. 6-7

Stoddard, Lothrop (1922) The New World of Islam, London, Chapman & Hall Ltd.

2nd impression

Stoianovich, Traian (1960) "The Conquering Balkan Orthodox Merchant" Journal of

Economic History, 20

Stoyanov, Iurii (1994) The Hidden Tradition in Europe: the Secret History of

Mediaeval Christian Heresy, Harmondsworth, Penguin

Arkana

Strand, Elin Pernilla (2001) Moving Hearts; Pentecostalism & Gypsy Identity,

unpublished MA thesis, London, University of Greenwich

Streck, Bernhard (2004) "Local strangers: the Nile Valley Gypsies in the ethnic

mosaic of the Sudan", in Aparna Rao & Joseph C. Berland

[eds. ], Customary Strangers: New Perspectives on

Peripatetic Peoples in the Middle East, Africa and Asia,

Westport USA, Praeger Publishers, pp. 179-94

teveenko, Nancy P. (2002) "Wild Animals in the Byzantine Park", in Henry Maguire,

Anthony Littlewood & Joachim Wolschke-Bulmahn [eds. ]

Byzantine Garden Culture, Washington D. C. Dombarton

Oaks Research Library & Collection, pp. 70-86

Hubel, Daniel (2006) "Myth, collective trauma and war in Serbia: a cultural- hermeneutical appraisal", Anthropology Matters Journal,

vol. 8, no. 1, pp. 1-9,

http: //www. anthropologymatters. com/journal/2006- 1 /index. htm

Svanberg, Ingvar (1982) A Bibliography of the Turkish-Speaking Tribal Yürüks,

Stockholm, Uppsala University Press

Svensson, Birgitta (1993) Bortom all ära och redlighet : tattarnas spel med rättvisan,

Stockholm, Nordiska museets handlinger, Nordiska museet

Szädeczky-Kardoss, S. (1990) "The Avars" in Denis Sinor [ed. ] Cambridge History of Early Inner Asia, Cambs. Cambridge University Press, pp. 206-228

Tebbutt, Susan (1998) Sinti and Roma: Gypsies in German-Speaking Society and Literature, (Culture & Society in Germany 2), New York &

Oxford, Berghahn Books,

Tekin, Gönül Alpay (1995) "`Othmänli - Literature" C. E. Bosworth [et at eds. ] The

Encyclopaedia of Islam New Edition, Leiden, E. J. Brill

Terdiman, Richard (1989) "Is there Class in this Class? " in H. Aram Veeser [ed. ] The

New Historicism, London, Routledge, pp. 225-230

Tha'älibi, Abü Mansur (1968) The Latä'if al-ma'räif of Tha'älibi; the Book of Curious and

Entertaining Information, C. E. Bosworth [trans. &

introduction, with notes], Edinburgh, The University Press

at Edinburgh

Thapar, Romila (2000a) Narratives and the Making of History: Two Lectures,

Oxford India Paperbacks, New Delhi, Oxford University

Press

Thapar, Romila (2000b) "Hindutva and History; Why do Hindutva ideologues keep

flogging a dead horse? " Frontline, vol. 17, no. 20, October

13, pp. 15-16

Theophanes (1982) The Chronicle of Theophanes : An English Translation of

anni mundi 6095-6305 (A. D. 602-813), Harry Turtledove

[trans], Philadelphia, University of Pennsylvania Press

Thubron, Colin (1994) The Lost Heart of Asia, London, Verso

Todorov, Nikolai (1983) The Balkan City, 1400 - 1900, Seattle & London,

University of Washington Press

Todorova, Maria (1997) Imagining the Balkans, New York, OUP

Todorova, Maria (1996) "The Ottoman Legacy in the Balkans" in L. Carl Brown

[ed. ] Imperial Legacy: the Ottoman imprint on the Balkans

and the Middle East, New York, Columbia University Press

Togan, Isenbike (1992) "Ottoman history by Inner Asian norms" in Halil Berktay &

Suraiya Faroqhi [eds. ] New Approaches to State & Peasant

in Ottoman History, Library of Peasant Studies 10, London,

Frank Cass, pp. 185-210

Tomkinson, John L. (2004) Haunted Greece: Nymphs, Vampires and other Exotika,

Athens, Anagnosis

Treadgold, Warren (1995) Byzantium & Its Army, 284-1081, Stanford, California,

Stanford University Press

Treadgold, Warren (1979) "The Revival of Byzantine Learning and the Revival of the

Byzantine State", The American Historical Review, vol. 84,

no. 5 (December), pp. 1245-1266

Turner, Sir Ralph Lilley (1927) "The Position of Romani in Indo-Aryan: a reply to Dr J.

Sampson", Journal of the Gypsy Lore Society [3rd Series],

vol. 6, pp. 129-38

Turner, Sir Ralph Lilley (1926) "The Position of Romani in Indo-Aryan", Journal of the

Gypsy Lore Society [3rd Series], vol. 5, pp. 145-189;

reprinted in Collected Papers 1912-1973 (1975), London

and Oxford, Oxford University Press for the School of

Oriental & African Studies, pp. 251-290.

Ubicini, Jean Henri A. (1856) Letters on Turkey : an account of the religious, political,

.., social and commercial condition of the Ottoman Empire.

[trans. From the French of A. Ubicini] Lady Easthope,

London, John Murray

Vaihle, S. (1913) "Mopsuestia", Catholic Encyclopaedia, [multiple editors]

under the imprimatur of the Archbishop of New York, John

Cardinal Farley, New York, Robert Appleton Co.

van de Port, Mattijas (1998) Gypsies, Wars and Other Instances of the Wild: Civilization

and Its Discontents in a Serbian Town, Netherlands,

Amsterdam University Press

Veeser, H. Aram (1989) "Introduction" in H. Aram Veeser [ed. ] The New

Historicism, London, Routledge, pp. ix-xvi

von Clausewitz, Carl P. G. (1908) On War, Col. J. J. Graham [trans. ], Col. F. N. Maude [new

& rev. ed. ], London, Trubner & Co., 3 vols.

von Hammer, Joseph (1854) History of the Ottoman Turks: from the beginning of their

empire to the present time; chiefly founded upon Von Hammer, 2 vols. [new edition 1856; new & revised editions 1858,1877,1878], London, Richard Bentley & Son

Voskanian, Vadan (2003) "The Iranian Loan-words in Lomavren, the Secret

Language of the Armenian Gypsies", Journal of Iran & the

Caucasus, vol. 6, no. 1-2, E. J. Brill, Leiden, pp. 169-80

Vryonis, Spyros Jr. (1975) "Nomadization and Islamization in Asia Minor" Dumbarton

Oaks Papers, 29, pp. 41-71

Vryonis, Spyros Jr. (1971) The Decline of Medieval Hellenism in Asia Minor and the

Process of Islamization from the Eleventh through the Fifteenth Century, [Publications of the Center for Medieval

and Renaissance Studies 4] Berkeley, California, UCLA

Vryonis, Spyros Jr. (1956) "Isidore Glabas and the Turkish devýirme", Speculum,

vol. 3 1, no. 3, pp. 432-455

Watt, W. Montgomery (2003) "al-Madina; History to 1926" in P. Bearman et al [eds] The

Encyclopaedia of Islam, WebCD Edition, Leiden, Brill, V,

994a-999a

Watt, W. Montgomery (1956) Muhammad at Medina, Oxford and New York, Oxford

University Press

Wayland Barber, Elizabeth (1999) The Mummies of Ürümchi, London, W. W. Norton & Co.

Wells, Peter S. (c. 1999) The Barbarians Speak; How Conquered Peoples Shaped

Roman Europe, Princeton NJ, Princeton University Press

Welch, S. C. (2003) "Arabiyya", in P. Bearman et al [eds] The Encyclopaedia of Islam, WebCD Edition, Leiden, E. J. Brill, vol.!, pp. 561-

603

White Jr. Lynn (1974) "Indic Elements in the Iconography of Petrarch's Trionfo

della morte", Speculum: A Journal of Mediaeval Studies,

vol. XLIX, no. 2, April, pp. 201 - 221

White, Hayden (1989) "New Historicism: A Comment" in H. Aram Veeser [ed. ]

The New Historicism, London & New York, Routledge,

pp. 293-302

White, Hayden (1987a) The Content of the Form: Narrative Discourse & Historical

Representation, Baltimore, John Hopkins University Press

White, Hayden (1987b) "The Value of Narrative in the Representation of Reality"

The Content of the Form: Narrative Discourse & Historical

Representation, Baltimore, John Hopkins University Press

White, Hayden (1978) Tropics of Discourse; Essays in Cultural Criticism,

Baltimore & London, John Hopkins University Press

White, Hayden (1976) "The Noble Savage Theme as Fetish" in F. Chiappelli [ed. ],

First Images of America: the Impact of the New World on the Old, Berkley & Los Angeles University of California

Press

White, Hayden (1976) "The Tropics of History: the Deep Structure of the `New

Science"' in G. Tagliacozzo & D. P. Verene [eds. ],

Giambattista Vico's Science of Humanity, Balimore &

London, John Hopkins University Press

White, Hayden (1975) Metahistory; The Historical Imagination in Nineteenth- Century Europe, Baltimore, John Hopkins University Press

White, Hayden (1974) "The Historical Text as Literary Artifact", Clio, 3/3 White, Hayden (1972) "The Forms of Wildness... " in E. Dudley & M. E. Novak

[eds. ], The Wild Man Within: An Image in Western Thought from Renaissance to Romanticism, Pittsburgh, Pittsburgh University Press

White, Hayden (1972) "The Irrational and the Problem of Historical Knowledge... " in H. E. Pagliaro [ed. ], Irrationalism in the

Eighteenth Century; Papers presented at the Second Annual Meeting of the American Society for Eighteenth Century

Studies, Studies in Eighteenth-Century Culture, Volume 2,

Cleveland

White, Hayden (1966) "The Burden of History", History & Theory, vol. 5, no. 2

White, Karin (2000) "Metal-workers, agriculturists, acrobats, military-people

and fortune-tellers: Roma (Gypsies) in and around the

Byzantine Empire", Golden Horn/Gouden Hoorn, vol. 7,

no. 2, http: //tinyprl. com/4knc7n

Wickens, G. M. (1976) "Persian literature: an affirmation of identity" in R. M.

Savory [ed. ] Introduction to Islamic Civilisation,


Willems, Wim (2003) "The Weakness of Well Ordered Societies: Gypsies in

Western Europe, the Ottoman Empire, and India, 1400-

1914" [with Leo Lucassen], Review, XXVI, no. 3,2003,

pp. 283-313

Willems, Wim (2000) "Gypsies in the diaspora? The pitfalls of a biblical concept"

[with Leo Lucassen] Histoire Sociale/Social History, vol. 33, no. 6, pp. 251-269

Willems, Wim (1998) Gypsies & Other Itinerant Groups: A Socio-Historical

Approach, [with Leo Lucassen & Annemarie Cottaar]

London & New York, St. Martin's Press

Willems, Wim (1993) In Search of the True Gypsy: from Enlightenment to Final

Solution, Don Bloch [trans. ] London, Frank Cass

Windfuhr, Gernot L. (2003) "Gypsy Dialects" in Ehsan Yarshater [ed. ] Encyclopaedia

Iranica, New York, Encyclopaedia Iranica Foundation, Vol.

XI, pp. 415a-421b

Ye'Or, Bat (2001) "Contemporary Arab & Muslim perceptions of the Other. "

The Contemporary Middle East: minorities, pluralism, Dhimmitude, Presentation at the International Conference

at the Simon Wiesenthal Centre, Paris

Ye'Or, Bat (1996a) The Decline of Eastern Christianity under Islam: from Jihad

to Dhimmitude, 7th-20th Centuries, Miriam Kochan and David Littman [translated from the French], Cranbury, NJ,

Fairleigh Dickinson University Press/Associated University

Presses

Ye'Or, Bat (1996b) "The decline of Eastern Christian communities in the

modern Middle East" Lecture at the Hebrew University of Jerusalem, Middle Eastern Unit & Faculty of Humanities

Ye'Or, Bat (1995) "Myth & Politics: origin of the myth of a tolerant pluralistic

I4SS19390 - Greenwich Academic Literature Archive (GALA)

Documents

Transcript of I4SS19390 - Greenwich Academic Literature Archive (GALA)